Integrated high-density genetic maps in multi-parent F(2) mapping populations provide a framework to unravel the genomic basis of agro-qualitative and metabolic traits in pepper (Capsicum annuum).

Integrated high-density genetic maps in multi-parent F2 mapping populations provide a framework to unravel the genomic basis of agro-qualitative and metabolic traits in pepper (Capsicum annuum) - PMC Skip to main content An official website of the United States government Here's how you know Here's how you know Official websites use .gov A .gov website belongs to an official government organization in the United States. Secure .gov websites use HTTPS A lock ( Lock Locked padlock icon ) or https:// means you've safely connected to the .gov website. Share sensitive information only on official, secure websites. Search Log in Dashboard Publications Account settings Log out Search… Search NCBI Primary site navigation Search Logged in as: Dashboard Publications Account settings Log in Search PMC Full-Text Archive Search in PMC Journal List User Guide PERMALINK Copy As a library, NLM provides access to scientific literature. Inclusion in an NLM database does not imply endorsement of, or agreement with, the contents by NLM or the National Institutes of Health. Learn more: PMC Disclaimer | PMC Copyright Notice BMC Plant Biol . 2026 Mar 9;26:684. doi: 10.1186/s12870-026-08517-7 Search in PMC Search in PubMed View in NLM Catalog Add to search Integrated high-density genetic maps in multi-parent F 2 mapping populations provide a framework to unravel the genomic basis of agro-qualitative and metabolic traits in pepper ( Capsicum annuum ) Annalisa Cocozza Annalisa Cocozza 1 CREA Research Centre for Vegetable and Ornamental Crops, Via dei Cavalleggeri 51, Pontecagnano Faiano, SA 84098 Italy 2 Department of Agricultural Sciences, University of Naples Federico II, Portici, NA Italy Find articles by Annalisa Cocozza 1, 2 , Antonietta D’Alessandro Antonietta D’Alessandro 1 CREA Research Centre for Vegetable and Ornamental Crops, Via dei Cavalleggeri 51, Pontecagnano Faiano, SA 84098 Italy Find articles by Antonietta D’Alessandro 1 , Rosaria Macellaro Rosaria Macellaro 1 CREA Research Centre for Vegetable and Ornamental Crops, Via dei Cavalleggeri 51, Pontecagnano Faiano, SA 84098 Italy Find articles by Rosaria Macellaro 1 , Gianluca Francese Gianluca Francese 1 CREA Research Centre for Vegetable and Ornamental Crops, Via dei Cavalleggeri 51, Pontecagnano Faiano, SA 84098 Italy Find articles by Gianluca Francese 1 , Pasquale Tripodi Pasquale Tripodi 1 CREA Research Centre for Vegetable and Ornamental Crops, Via dei Cavalleggeri 51, Pontecagnano Faiano, SA 84098 Italy Find articles by Pasquale Tripodi 1, ✉ Author information Article notes Copyright and License information 1 CREA Research Centre for Vegetable and Ornamental Crops, Via dei Cavalleggeri 51, Pontecagnano Faiano, SA 84098 Italy 2 Department of Agricultural Sciences, University of Naples Federico II, Portici, NA Italy ✉ Corresponding author. Received 2026 Jan 8; Accepted 2026 Mar 3; Collection date 2026. © The Author(s) 2026 Open Access This article is licensed under a Creative Commons Attribution-NonCommercial-NoDerivatives 4.0 International License, which permits any non-commercial use, sharing, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons licence, and indicate if you modified the licensed material. You do not have permission under this licence to share adapted material derived from this article or parts of it. The images or other third party material in this article are included in the article’s Creative Commons licence, unless indicated otherwise in a credit line to the material. If material is not included in the article’s Creative Commons licence and your intended use is not permitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder. To view a copy of this licence, visit http://creativecommons.org/licenses/by-nc-nd/4.0/ . PMC Copyright notice PMCID: PMC13085554  PMID: 41803704 Abstract Background Pepper ( Capsicum spp.) is a major vegetable crops renowned for its nutritional value and for the content of bioactive compounds with beneficial health effects. Due to its extensive agro-qualitative variability, this crop has a wide range of food and non-food uses, which drive its commercial value. In this study, quantitative trait loci (QTL) mapping and candidate gene analyses were used dissect the genetic architecture of key traits impacting overall pepper fruit quality. Results Two intra-specific F 2 populations ( n = 226) were developed by crossing highly phenotypically divergent and genetically distant parents. Using double digest restriction-site associated DNA (ddRAD-seq), over 5,000 non-distorted SNP markers were used to construct two high density genetic maps spanning approximately 2.500 cM and defining over 4,500 bins. Metabolomic profiling focused on three classes of primary and secondary metabolites including sugars (glucose, fructose, sucrose), organic acids (quinic, malic, succinic, citric, and ascorbic acid) and carotenoids (β-carotene). Simultaneously, 45 fruit agro-morphological and colour traits were assessed through automated phenotyping tools and digital analysis of over 2500 fruit section. In total, 200 QTLs were detected via inclusive composite interval mapping, with 113 QTLs co-localizing into 23 pleiotropic clusters. Eighteen candidate genes involved in the regulation of fruit organization, differentiation, and associated metabolic pathways were identified within QTL peak regions. Notably, a major hotspot at the base of chromosome 3 was found to coordinate increased sugar content and fruit size with decreased citric acid. Additional pleiotropic QTLs on chromosomes 5, 9, and 12 simultaneously increased malic acid, sucrose, and citric acid alongside fruit size parameters. Overall, agronomic and morphological traits followed a predominantly additive model, whereas many metabolite-related traits exhibited dominance or overdominance. Conclusions The work aims to bridge the gap in the limited research undertaken in Capsicum to incorporate QTL analysis for agronomic, morphological, and metabolite aspects fundamental for determining the organoleptic and market-related characteristics of the fruit. Indeed, the genetic basis of many of these traits remains still largely unknown. Results from this study provide novel insight to define roadmaps toward genetic improvement for quality in pepper. Supplementary Information The online version contains supplementary material available at 10.1186/s12870-026-08517-7. Keywords: Pepper, F 2 , DdRAD-seq, Genetic maps, QTL mapping, Bioactive compounds, Digital phenotyping, Qualitative traits, Fruit size and shape Introduction As staple vegetable at the base of the food pyramid, pepper ( Capsicum spp.) contributes significantly to human dietary balance by providing a wealth source of bioactive compounds. The genus, part of the large Solanaceae family, had its ancient origin in a wide area including Central and South American countries, where over 40 species were domesticated and grown as food crops and for medical relief [ 1 , 2 ]. In the post-domestication period and following trans-continental trades during the Age of discovery, it has been spread to Africa, Asia, and Europe, where different sweet and hot types have been developed to suit diverse climatic environments and culinary traditions [ 3 ]. Nowadays, the crop is economically relevant, ranking amongst the first top seven in terms of world total production being cultivated in both tropical and temperate zones on an area of approximately 4 million hectares with a total output of 42 million tons (Chillies and peppers, green and dry) [ 4 ]. Over the last three decades, the main objectives of pepper genetic improvement have focused on the selection of high-yielding varieties with good environmental adaptability and resistance to various pests and diseases [ 5 , 6 ]. Although these goals remain priorities, shifting market preferences changes and customers’ awareness of healthy products require increased effort in breeding for quality. Indeed, modern consumers pay growing attention to the visual appearance of fruits [ 7 ], but also to the nutraceutical properties of horticultural products [ 8 ]. In this context, global fruit quality plays a strategic role in breeding programs, no longer limited to resistance or productivity, but also oriented towards the enhancement of the metabolic and sensory components. In pepper, quality is the combination of all agro-qualitative characteristics defining its product and nutritional properties and are associated to size, shape, color, weight and metabolite content of fruits. Fruit shape and size are the main traits underpinning the marketable value; typically, blocky and trapezoidal fruits are typically associated with sweet flavours; elongated and tapered fruits are preferred for spice production and associated with pungent aroma, while small cherry-like fruits are desirable for ornamental use [ 9 , 10 ]. The different typologies can also be adapted to different market destinations, from fresh consumption to processed products such as powders, pastes, or sauces [ 9 ]. Furthermore, agro-qualitative characteristics are closely linked to the content of bioactive compounds; for instance, the colour of the fruit at the ripening stage is due to the accumulation of carotenoids (capsanthin, capsorubin, β-carotene), which also contribute to the pepper’s nutritional value [ 11 ]. Once ripened, fruits are also richer in vitamin C, a water-soluble metabolite that in pepper may reach amounts three to five times the recommended daily intake for a balanced human diet [ 12 , 13 ]. These compounds exert significant antioxidant action as free radical scavengers to neutralize the detrimental effects of reactive oxygen species (ROS), thus providing inflammatory effects and preventing several chronic diseases [ 14 ]. The dietary intake of these substances is fundamental since both β-carotene and Vitamin C cannot be synthesized in the human body [ 14 , 15 ]. Pepper fruit quality is also strongly influenced by the concentration of soluble sugars and organic acids, playing a central role both for the physiology of the plant and for the nutritional and functional value of the fruit. The flavour and taste of pepper fruits are mainly determined by the soluble sugar and organic acid content as the main sensory parameters regulating the perceived sweetness and sourness [ 16 ]. The biological complexity of these quality traits is typically regulated by multiple genes, necessitating genome-mediated breeding strategies to unravel their genetic architecture. Essential requirements for mapping traits having a quantitative genetic inheritance, known as QTLs (quantitative trait loci), are the development of recombinant mapping populations and the establishment of dense genetic maps [ 17 ]. Several methods for QTL analysis have been proposed, including the simple regression (SM) and interval mapping (IM) models [ 18 ]. While SM identifies individual markers associated with a phenotype, IM evaluates flanking markers simultaneously. However, both approaches hold several drawbacks: SM ignores associations between linked markers, potentially underestimating QTL effects, while IM may miss QTLs outside the specific interval. Therefore, to increase the statistical power and efficiency, methods combining the strengths of both models, such as composite interval mapping (CIM) or inclusive composite interval mapping (ICIM), have been proposed [ 19 , 20 ]. The effectiveness of these mapping models is increased by the capacity to develop genetic maps with thousands of single nucleotide polymorphisms (SNP) markers [ 21 ]. In pepper, several genetic maps have been created at the inter- and intra-specific level in biparental F 2 populations through PCR-based markers [ 22 ]. These were mainly used to dissect the genetic basis of resistance traits, and although provided a limited coverage at the genome level, they also served to identify main regions underpinning the variation of the weight and shape of the fruits on the chromosomes 2, 3, and 10 [ 23 – 27 ]. Ultra dense genetic maps have been constructed after the release of the first reference genome sequence and the implementation of new genomic technologies [ 28 , 29 ] that promoted a higher resolution scan of genomic regions underlying QTLs. These have been useful for identifying QTL of agronomic interest, for instance, eighty-six QTLs for plant architecture and fruit traits were discovered by a low-coverage resequencing-based genetic map spanning 1372 centimorgan (cM) [ 28 ] while 12 QTLs for fruit weight and fruit length were found throught a GBS-based genetic map of 1,123.6 cM [ 30 ]. Despite confirming earlier findings, these studies revealed diverse genomic regions controlling the variation in fruit weight and shape traits, emphasizing the need to more accurately dissect the pepper genome. As for metabolites, most studied have been addressed at capsaicinoids while only recently the QTLs for primary and secondary metabolites have been mapped combining linkage and genome wide association approaches (GWAS) [ 31 , 32 ], shedding light on the complex genetic architecture of these traits. Therefore, it is necessary to integrate further studies in different genetic backgrounds to validate and identify new genomic regions associated with key metabolites. In the present study, QTL analysis for fruit quality related traits was performed in two intraspecific F 2 populations derived from highly diverse founder lines pepper. Double-digest restriction-site associated DNA sequencing (ddRAD-seq) was used for genotyping the two populations and constructing the genetic maps. By reducing genome complexity, this method effectively captures a high number of genome wide SNPs thus providing high-density genetic maps [ 33 , 34 ]. To gain insight into the genetic mechanisms governing fruit agro-quality traits in pepper, nine biochemical compounds—including primary sugars, organic acid, beta carotene, and vitamin C—as well as forty-five features related to fruit weight, color, size and shape using digital imaging, were taken into consideration. A better understanding of the genetic basis of agro-qualitative traits will provide powerful tools for breeders to develop superior cultivars, enhancing both economic value and consumer acceptance. Materials and methods Plant material and growth conditions Three C. annuum cultivars were used as founder lines to develop the intraspecific F 2 mapping populations: Early California Wonder (ECW), a European elite sweet pepper, with yellow, blocky fruits; Serrano Criollo de Morelos cv. CM334 (SCM), a pungent, red, horn-shaped Mexican landrace; Perennial (PER), a highly pungent Indian variety with small red fruits. The three parent lines were chosen for their high phenotypic, genetic and geographical divergence, and for being recognized for their use in genetic pepper studies [ 25 , 28 , 35 ]. The first population was derived from the cross between ECW as female parent and SCM as male parent, while the second resulted from a cross between PER as female parent and ECW as male. Hybridization was conducted in a growth chamber, and F 1 progenies were validated with molecular markers to determine success of the hybrid combination. Hybrids were then self-crossed to obtain the F 2 generation. A total of 226 F 2 individuals (100 ECW x SCM and 126 PER x ECW) alongside with respective parent lines were grown from April to October 2023 in the greenhouses of CREA Research Centre for Vegetable and Ornamental Crops (CREA OF), Pontecagnano, Salerno, Italy Crops (40°65′ N; 14°89’ E) (Fig. 1 ). The internal temperature (°C), relative (°RH) and absolute (g kg − 1 ) humidity and the external radiation (w m − 2 ) were recorded. An automated climate control system consisting of an evaporative fan cooling system and a shading screen were adopted, ensuring temperatures during the growth cycle not exceeding 30 °C. Standard procedures for managing pests and diseases were followed to guarantee consistent plant growth. Fig. 1. Open in a new tab Experimental populations used to identify the genetic basis and agro-quality traits of peppers. On the top are indicated the two founder lines used to develop the respective populations Agro qualitative trait detection For each F 2 individual, five mature fruits were harvested, washed and individually evaluated for agro-qualitative traits. Proximate traits included fruit weight (FW), fruit length (FL) and fruit width (FD). Fruit colour was assessed through a CR-210 Chroma Meter (Minolta Corp., Osaka, Japan). Measurements were performed at the midpoint between the distal and the basal ends of the fruit and expressed as CIELAB L*, a*, and b* values. L* indicates lightness/darkness (0 = black, 100 = white), a* describes intensity in green − red (where a positive number indicates redness and a negative number indicates greenness), and b* describes the intensity in blue−yellow (where a positive number indicates yellowness and a negative number indicates blueness). The chroma value indicating colour saturation was estimated from a* and b* using the formula: [(a*) 2 + (b*) 2 ] 0.5 . Digital analysis of fruit size and shape traits Fruit size and shape parameters were obtained through digital image analysis. For each F₂ individual and parental line, five representative fruits were evaluated. This sampling strategy ensured adequate representation of genotype variation and provided a robust dataset for quantitative morphological analysis. Each fruit was longitudinally sectioned into two halves, and the resulting fruit sections were scanned at a resolution 300 dpi using a CanoScan LiDE 210 photo scanner (Canon, Tokyo, Japan). Fruit scans were then processed using the Tomato Analyzer v 3.0 software [ 36 ]. Thirty-eight quantitative descriptors, categorized into: fruit size (7), shape index (3), blockiness (3), homogeneity (3), proximal fruit end-shape (4), distal fruit end shape (4), asymmetry (6), internal eccentricity (5), latitudinal Sect. (3), were automatically recorded. Manual adjustments were performed when the software failed to accurately identify the fruit outline. Biochemical analysis For metabolite compounds detection, three biological replicates, each containing a 20-grams pooled fruit sample from individual plants, were analyzed. Sampled fruits were deprived of peduncle, washed, and dried. After removing the seeds, fruits were longitudinally halved, flash-frozen in liquid nitrogen, and stored at −80 °C. For each pooled sample, two technical replicates were extracted and analyzed via HPLC (High-performance liquid chromatography). Ascorbic acid and beta carotene determination The ascorbic acid content (AsA) was measured by mixing 1 g of the pepper sample with 3 ml of 6% metaphosphoric acid in distilled water. The mixture was blended for 30 s using an Ultra-Turrax homogeniser (IKA, Wilmington, NC, USA) and centrifuged at 2,000 g for 15 min. This process was repeated twice, and the supernatants were pooled and filtered through a 0.22 μm polytetrafluoroethylene (PTFE) filter. Analysis was performed in an Ultimate 3000 UPLC system (Thermo Fisher Scientific, Sunnyvale, CA, USA) equipped with a Kinetex C18 column (75 mm x 4.6 mm, 100 Å, particle size 2.6 μm; Phenomenex) using 5 ml sample. The mobile phase consisted of 0.02 mol L − 1 phosphoric acid at a flow rate of 0.35 mL min − 1 . AsA was detected at 254 nm and quantified using a calibration curve of pure standard (Sigma-Aldrich). For beta-carotene extraction, 5 g of the sample were mixed with 15 ml of buffer solution (methanol + BHT + EDTA), homogenised and filtered with a PTFE membrane. The filtrate was subjected to double extraction with 20 mL of dichloromethane, followed by evaporation to dryness at 35 °C using a rotary evaporator. The residue was reconstituted in 1 mL of tetrahydrofuran and vortexed. Quantification was performed using a β-carotene standard (Sigma-Aldrich, St. Louis, MO, USA). Sugars and organic acids composition Soluble sugars (glucose, fructose, and sucrose) were extracted from 100 mg of frozen tissue using 1 mL of 80% ethanol for 30 min. Samples were centrifuged at 15,000 g for 10 min, and the resulting supernatants were filtered through a PTFE membrane. Analysis was performed by injecting a 20 µL aliquot onto a Luna NH₂ 100 Å column (250 × 4.6 mm, 5 μm particle size) equipped with a SecurityGuard NH₂ guard column (4 × 3.0 mm); both were supplied by Phenomenex (Torrance, CA, USA). The mobile phase consisted of a 75:25 (v/v) mixture of acetonitrile and water, with a flow rate of 1 ml/min. Organic acids (citrate, malate, quinate, and succinate) were extracted by diluting 5 mg of sample in 2 mL of Milli-Q water, followed by homogenization and centrifugation at 14,000 g for 20 min. Analysis was performed using a Waters E-Alliance HPLC system equipped with a 2695 separation module and a Model 410 refractive index detector (Waters Corporation, USA). Separation was achieved on a Rezex ROA-Organic Acid H⁺ (8%) column (150 × 7.8 mm; Phenomenex, Torrance, CA, USA) using an isocratic mobile phase of 1 mM H₂SO₄ at a flow rate of 0.5 mL min − 1 [ 37 ]. Peak areas were recorded and concentrations (g 100 g − 1 FW for sugars and mg 100 g − 1 FW for organic acids) were quantified using authentic standards (Sigma-Aldrich, St. Louis, MO, USA). Data acquisition was managed through Empower software (Waters Corporation). ddRad-seq genotyping Young leaves of parents and F 2 progenies were collected and quickly frozen in dry ice prior lyophilization. Lyophilized tissue was grind using the Tissuel Lyser II (Qiagen, Hilden, Germany) and genomic DNA isolated using the DNeasy ® Plant Mini Kit (Qiagen, Hilden, Germany) following the manufacturer’s instructions. DNA concentration and quality were measured by absorbance at 260 and 280 nm, respectively, using both a UV-Vis spectrophotometer (ND-1000; NanoDrop, Thermo Scientific, Wilmington, DE, USA) and a Qubit 2.0 Fluorometer based on the Qubit dsDNA HS Assay (Thermo Fisher Scientific, Waltham, MA, USA). Double-Digest Restriction-Site-Associated DNA Sequencing (ddRad-seq) was performed with 2 µg of genomic DNA digested with Mbo I and Sph I (New England BioLabs, Ipswich, MA, USA) endonucleases following the IGA technology services s.r.l (Udine, Italy) protocol. Briefly, fragmented DNA was ligated to barcoded adapters, pooled on multiplexing batches, and purified with AMPure XP beads (Agencourt, Beverly, MA, USA). For each pool, targeted fragments were size selected using a BluePippin instrument (Sage Science Inc., Beverly, MA, USA). The gel eluted fraction was amplified with indexed primers and products purified with AMPureXP beads. The resulting libraries were validated using both Qubit 2.0 Fluorometer (Invitrogen, Carlsbad, CA, USA) and Bioanalyzer DNA assay (Agilent technologies, Santa Clara, CA, USA). Sequencing was performed on an Illumina NovaSeq 6000 platform (Illumina, San Diego, CA, USA). in 150-bp paired-end mode. Genomic analysis and genetic map construction Illumina reads were demultiplexed using the process_radtags utility included in the Stacks v2.0 software (Rochette et al., 2019). Reads were aligned to the Capsicum annuum var. Zhangshugang reference genome [ 38 ] using the Burrows–Wheeler Aligner BWA-MEM v0.7.17 [ 39 ] with default parameters. Detection of all the covered loci from the aligned reads was performed using the gstacks program included in Stacks v2.61 [ 40 ]. Filtering of detected loci was done with the populations program included in Stacks v2.61 applying the following criteria: loci were retained if represented in at least 75% of the individuals (– R = 0.75) and exhibited a maximum observed heterozygosity of 80% (--max-obs-het 0.8) to exclude potential paralogs and misaligned reads. To obtain markers for genetic map construction and QTL analysis, subsequent filtering steps were applied: (a) application of a Minor Allele Frequency (MAF) of 0.05, (b) removal of monomorphic and redundant markers in both populations, (c) removal of heterozygous markers in the parental lines, (d) deletion of distorted markers departing from the expected 1:2:1 segregation ratio ( P < 0.001). A total of 5,176 markers in the ECW x SCM population and 5,206 markers in the PER x ECW population were retained. The two genetic maps were constructed with JoinMap ® 4.0 software [ 41 ]. A minimum LOD threshold of 7.0 with a recombination frequency of 0.45 were chosen for mapping. The regression mapping algorithm was used to build the linkage map, and the Kosambi mapping function was utilized to convert the recombination frequencies into map distance in centiMorgans (cM). Genetic and physical maps were compared with the genetic map comparator [ 42 ]. Colinearity between genetic and physical positions of the mapped markers was determined by producing scatter plots of the female and male marker distribution, plotted using ggplot2 [ 43 ]. The integrated map was constructed using JoinMap ® 4.0 using the regression mapping algorithm. Statistical data analysis All data were analysed using the R environment (version 4.1.2; R Core Team, 2021). One-way analysis of variance (ANOVA) was performed to evaluate differences among genotypes, and mean separation was performed using the Tukey honest significance difference (HSD) test at p ≤ 0.05 implemented with the ‘ multcompView ’ package in R. Relationships among phenotypic traits were assessed using Pearson’s correlation coefficients. Correlation networks were visualised using Cytoscape [ 44 ] and further explored using Metascape tools [ 45 ] to highlight relevant trait associations and clusters. Heatmap for metabolites was permormed using MetaboAnalyst software (version 6.0). All plots were drawn using the R packages ‘ ggplot2 ’, ‘ ggbiplot ’ and ‘ heatmap ’, QTL mapping analysis Prior to QTL analysis, all phenotypic data were reviewed for outliers using the 1.5 × IQR (interquartile range) method. For metabolic traits, technical replicates were averaged to obtain a mean value per biological replicate, which were then combined into a single phenotypic value for each F 2 individual. Similarly, measurements from individual fruits for agronomic and morphometric traits were averaged to provide a representative value per plant. The number of individuals per population and the specific traits analyzed in each category are summarized in Table 1 . Table 1. Summary of trait categories, number of traits analyzed, and population sizes (N) used for QTL mapping in the ECW x SCM and PER x ECW F 2 populations Trait Category Traits Number of traits Population size (ECW x SCM) Population size (PER x ECW) Primary Metabolites Sugars 3 100 126 Secondary Metabolites Organic acids and carotenoids 6 100 126 Agronomic Fruit weight, Fruit length, Fruit width, CIELAB coordinates 7 100 126 Digital Imaging Size Fruit size 7 100 126 Digital Imaging shape Shape index, Blockiness, Homogeneity, proximal and distal fruit end-shape, asymmetry, internal eccentricity, latitudinal section 31 100 126 Open in a new tab QTL mapping was conducted using the inclusive composite interval mapping (ICIM) method implemented in the QTL IciMapping 4.1. software [ 46 ]. The ICIM approach combines two-mapping strategies: first, a stepwise regression equation identifies the most significant markers for QTL mapping; then, interval mapping is carried out using phenotypic values adjusted by the marker variables kept in the first step. With this model, the modified phenotype includes the position and additive effect data of the QTL, excluding any background effects of the QTL and/or genetic variations on other regions of the genome, thus increasing the accuracy and power of detection and reducing bias in the estimation of QTL’s effects [ 47 ]. A likelihood of odds (LOD) threshold above 3.0 was applied as the minimum significance level for the main-effect QTLs, implementing stepwise regression method with a 0.5 cM window size and probability levels for entering and removing variables (PIN) set at 0.01. This threshold was considered conservative given the stringent background control of the ICIM algorithm, which utilizes stepwise regression to minimize false positives and ‘ghost’ signals [ 46 ], ensuring a balanced control between Type I and Type II errors across the diverse set of traits analyzed. The QTL confidence intervals were determined using the one-LOD drop criterion, defined as the genomic region where the LOD score remains within one unit of its peak value. QTL clusters were identified based on overlapping confidence intervals on the genetic map. Shared QTLs across populations were identified based on overlapping genetic intervals on the integrated map. For each QTL is shown the peak and interval regions in megabase pair (Mbp) and centimorgan (cM), the phenotypic variation explained in percentage (PVE%), the additive and dominant effects and the genetic mode of action. Each QTL was named according to standard rules that consider the acronym of the trait, the reference chromosome and, after a dot, a digit indicating a progressive number within the same chromosome (e.g. qL4.1 refers to q = qtl, L = L* color coordinate, 4 = chromosome 4, 0.1 = first QTL found for the trait in the chromosome). Candidate genes within the identified QTL intervals were prioritized based on their proximity to the peak marker and functional congruency with the trait’s biology, as supported by prior literature in the same or related species. Physical mapping of significantly associated SNPs and functional annotation of the predicted underlying genes were performed using the C. annuum var. Zhangshugang gene models [ 48 ] and further refined with InterProScan [ 49 ]. For SNPs mapping within intronic regions, we considered the nearest genes located upstream or downstream of the significant markers. Genes linked to the pepper-associated traits have been proposed as potential causative genes and their function compared to the literature and to the UniProt database [ 50 ]. Results Agronomic trait variation across parental’s lines and in F 2 populations Agronomic and colorimetric parameters (FW, FL, FWD, L*, a*, b*, and Chroma) are summarized in Table 2 . FW showed wide variation in both populations, with higher mean values in ECW x SCM (19.79 gr) compared to PER × ECW (10.50 gr). A similar trend was observed for FL and FWD; specifically, the ECW × SCM population produced, on average, longer (79.90 mm) and wider fruits (26.25 mm) than PER × ECW (58.74 mm and 20.63 mm, respectively). In addition, in ECW x SCM, skewness values ranged from − 0.51 to 2.24, indicating that most variables followed a symmetric distribution, with only a few showing a slight trend towards higher values. A similar pattern was observed in the PER × ECW population (skewness from − 0.34 to 2.39), where most traits remained nearly symmetrical, with only a few variables exhibiting more pronounced asymmetry. Table 2. Descriptive statistics of agronomic traits for each F₂ population and for parental genotypes. Different letters within each column indicate significant differences among genotypes according to Tukey’s HSD test ( p ≤ 0.05) Trait* SCM ECW PER ECW × SCM PER × ECW Mean Mean Mean Mean Range SD skew kurtosis Mean Range SD skew kurtosis FW (gr) 4.96 c 79.7 a 0.62 a 19.79 b 1.70–90.10 9.4 2.24 9.94 10.5 a 1.30–55.00 6.2 1.53 4.71 FL (mm) 53.65 ac 55.91 a−c 25.88 a 79.9 b 10.91–127.15.91.15 19 −0.03 −0.07 58.7 c 21.31–131.75.31.75 15 0.52 0.83 FWD (mm) 16.55 a 66.59 a 6.49 a 26.25 a 11.06–52.83 6.5 0.82 1.74 20.6 a 2.52–46.86 6.9 0.8 0.59 L* 38.99 a 55.88 a 38.12 a 41.94 a 29.55–73.13 11 0.68 −1.04 46.6 a 29.54–67.23 12 0.32 −1.71 a* 31.33 ab 2.3 c 32.81 ab 23.77 a 5.14–37.36 9.2 −0.51 −1.25 26.7 b 3.16–60.08 12 −0.34 −0.82 b* 20.77 a 42.88 a 16.84 a 28.42 a 9.77–57.75 17 0.54 −1.58 37 a 6.72–147.04.72.04 24 1.17 2.49 Chroma 37.62 a 42.94 a 36.88 a 40.76 a 21.99–60.80 9.7 0.36 −1.17 50 a 13.31–150.92.31.92 17 2.39 11.31 Open in a new tab * Fruit weight FW is expressed in grams, fruit length FL and fruit width FWD are expressed in mm In both populations, kurtosis values ranged from − 1.58 to 9.84 (ECW × SCM) and − 1.71 to 11.31 (PER × ECW), indicating that most distributions were nearly normal or platykurtic while only a few showing leptokurtic or pronounced distribution. In both populations, digital colorimetric traits displayed bimodal distributions, highlighting the coexistence, of two main groups associated with red and yellow fruit phenotypes (Supplementary Fig. 1). This trend reflects the segregation of the color character among the individuals analyzed. Conversely, agronomic traits (FW, FL, FWD) showed continuous and moderate asymmetric distributions, appearing more regular and closer to normality in the PER × ECW population, particularly for fruit size-related traits. Overall, in both F 2 populations, the mean values were intermediate with respect to the parental genotypes that showed a considerable phenotypic variability for all the agronomic traits considered. Metabolic profiling of F₂ populations The distribution curves of metabolites in the two F₂ populations showed differences in their respective metabolic profiles (Supplementary Fig. 2). Both populations exhibited wide and continuous variation for the assayed metabolites, indicative of quantitative inheritance, with symmetry varying depending on the specific compound analyzed. Metabolomic profiling confirmed these differences, highlighting partially distinct profiles between the two populations (Fig. 2 ), with sugars and organic acids being primarily responsible for the metabolic diversification observed. In general, the ECW x SCM population showed slightly lower mean levels for most metabolites than the PER × ECW one (Table 3 ). Among organic acids, ascorbate and citrate show higher mean values in PER x ECW (104.38 mg 100 g -1 FW and 411.02 mg 100 g -1 FW) than in ECW x SCM (78.63 mg 100 g -1 FW and 295.30 mg 100 g -1 FW). As for succinate, the two populations show similar mean values, while malate and quinate were among the most variable compounds (110.06 mg 100 g -1 FW in ECW x SCM). β-carotene, representative of secondary metabolites and closely associated with the nutritional quality of the fruit, showed slightly higher mean levels in the ECW x SCM population. Fig. 2. Open in a new tab The heatmap shows the relative distribution of nine metabolites analyzed in both populations. The color scale (−6 to + 6) reflects relative abundance: blue hues indicate low concentration, while red indicates high metabolite presence Table 3. Descriptive statistics of metabolic compound for each F₂ population and for parental genotypes. Different letters within each column indicate significant differences among genotypes according to Tukey’s HSD test ( p ≤ 0.05) Trait* SCM ECW PER ECW × SCM PER × ECW Mean Mean Mean Mean Range SD skew kurtosis Mean Range SD skew kurtosis Fructose 1.84 b 1.36 ab 0.25 a 1.4 b 0.31–2.53 0.45 −0.43 0.32 1.35 b 0.03–3.87 0.73 0.29 1.04 Glucose 1.46 b 0.31 ab 0.14 a 0.94 ab 0.19–1.73 0.38 −0.13 −0.54 1.03 b 0.13–3.63 0.61 0.81 1.15 Sucrose 0.15 a 0.04 a 0 a 0.05 a 0.01–0.43 0.08 1.99 3.7 0.07 a 0–0.91.91 0.13 2.84 10.53 Ascorbate 144.98 c 31.52 b 575.89 a 78.63 b 9–153.12.12 29.64 0.03 0.42 104.38 c 11.25–223.74.25.74 40.3 0.07 0.57 Citrate 330 a−c 221.34 bc 563.38 a 295.3 b 100–619.08.08 114.6 0.07 −0.26 411.02 ac 9.12–1156.24.12.24 161.9 0.95 2.48 Malate 1405 c 18.5 b 4521.14 a 250.56 b 18–1591.6.6 437.5 1.75 1.7 353.95 b 19.13–3442.6.13.6 540.8 2.55 8.46 Quinate. 2175 b 717.89 ab 108.52 a 1091.1 ab 80–2818.72.72 772.8 −0.02 −1.1 1008.61 ab 29–3136.16.16 781.7 0.49 −0.82 Succinate. 302.5 b 69.72 ab 200.54 ab 110.06 a 6–463.88.88 97.44 1.28 1.39 173.45 b 8.08–666.84.08.84 131 1.29 2.14 β-carotene 2080.54 c 5.67 b 4219.2 a 534.23 b 2–2845.2.2 579.2 1.34 1.77 491.74 b 1.9–4266 560.8 2.92 13.91 Open in a new tab * Fructose (FRU), Glucose (GLU) and Sucrose (SAC) are expressed as g 100 g -1 FW; Ascorbate (ASA), Malate (MAL), Citrate (CIT) and Quinate (QUI) are expressed as mg 100 g -1 FW; β-carotene (B-car) is expressed as µg 100 g -1 FW These results form the basis for the identification of QTLs associated with metabolic profiles and for the selection of genotypes with superior quality characteristics. Two-dimensional digital analysis for fruit size and shape traits High-throughput digital analysis of over 2,530 fruit section images enabled the precise extraction of thirty-eight morphometric traits related to fruit shape, proportions, symmetry, and distal/proximal morphology. Analysis of variance (ANOVA), followed by the Tukey’s test ( p < 0.05), showed a different degree of phenotypic divergence in the two F₂ populations analyzed (Supplementary Tables 1 and 2 ). Among the parental lines, the greatest differences were observed for size-related traits, such as perimeter and area, with the PER exhibiting significantly lower average values than the other two founders in this respect. In contrast, parameters such as distal indentation area, distal end protrusion, obovoid and pericarp thickness had similar values in the three parental lines. For most of the traits considered, the mean values of the two F₂ populations were generally intermediate with those of the respective parents. However, some traits deviated from this trend and showed values outside the parental range. This trend was particularly evident for those traits related to distal and proximal morphology, fruit angular measures and for some shape indices. In particular, the ECW × SCM population showed significantly higher mean values for the proximal angle micro (PMI) and proximal angle macro (PMA) compared to both parents. On the contrary, the PER x ECW population exhibited lower distal angle micro (DMI) and macro (DMA) values than either parent. The average coefficient of variation (CV) was 25.88 in the ECW x SCM cross and 28.26 in the PER x ECW F 2 ’s. For both populations, distal end protrusion (DEA) showed the highest CV, being 100.0 and 127.3 for ECW x SCM and PER x ECW, respectively. The minimum CV was recorded for proximal eccentricity (1.12 in ECW × SCM) and eccentricity (2.63 in PER × ECW). Both populations showed the same minimum values equal to zero for the distal indentation area (DIA) and obovoid (OB) traits. This extensive variability observed for morphometric traits provided a solid basis for subsequent QTL mapping investigation. Multivariate analysis: PCA and network correlation between traits Principal Component Analysis (PCA) performed on all variables across the F 2 populations, and their parents accounted for 37.1% of the total phenotypic variance within the first due dimensions. (Fig. 3 ). The first component (PC 1 ) explaining 25.8% of the total variance was positively correlated to fruit size traits including perimeter, area, maximum height, maximum width, width mid-height, and pericarp area and negatively correlated to pericarp thickness, distal macro angle, distal micro angle. The second component (PC 2 ), which explaining 11.3% of the total variance was positively correlated to fruit shape indices (fruit shape index I e II), circularity, and obovoid, while being negatively correlated to fruit weight, proximal macro angle, proximal indentation area traits. Most metabolites were positively correlated with PC 2 , except for fructose and quinate, which exhibited positive correlations with PC 1 . These traits were the primary factors discriminating against the two populations, whereas colorimetric variables had a lower impact with L*, b* and Chroma being negatively correlated to both PCs and a* displaying positive correlations to PC 2 . The broad ellipses observed for both F 2 populations indicate high levels of internal phenotypic diversity. Additionally, the partial overlap of the two segregating groups suggests shared phenotypic variability, consistent with their common parent (ECW). Fig. 3. Open in a new tab Principal component analysis. Loading plot of the first (PC1) and second (PC2) principal components showing the variation for 54 agronomic, morphometric and metabolic traits in the two F 2 mapping populations. a Founder lines and respective populations are represented by different coloured symbols as indicated in the legend. Ellipses show the 95% confidence regions for each group. b Scree plot indicating the explained variance for each principal component. c The biplot showing the contribution of the variables to the first two principal components. The arrows represent the original variables, coloured according to their contribution to the total variance (colour scale from blue = low contribution to red = high contribution) Correlation network analysis (Fig. 4 ) revealed a predominance of positive correlations in both populations, mostly within the same trait categories (56.38% in ECW x SCM and 53.51% in PER x ECW). A slightly higher number of significant correlations were found in the ECW x SCM population. Fruit shape and size were strongly correlated, with a prevalence of positive correlations. Strong intra-category positive correlations were observed between the fruit shape variables (FSI, FSEI, FSEII, FSC, OV, C, LD and OV), many of which were negatively correlated with the width at mid-height (WMH) trait. Significant negative correlations were also evidenced between pericarp thickness and several size and shape traits. Among metabolic traits, malate and quinate displayed a high number of positive correlations with fruit shape traits. Notably, in the PER x ECW population a higher number of negative correlations were observed between metabolites and agronomic traits. In particular, ascorbic acid (ASA) was negatively correlations with all fruit size traits, FW and FL. Additionally, significant negative correlations were found within Obvoid and OC, Asov, FST, DFB, PFB and PEC. Fig. 4. Open in a new tab Correlation networks between morphological, agronomic and metabolic traits. The blue lines represent negative correlations, while the red lines indicate positive correlations; the thickness of the link indicates the intensity of the correlation. Only significant correlations ( P < 0.05) with coefficients in the range − 1.00 < −0.2 and 0.2 < 1.0 are displayed. The nodes are divided by category: fruit shape traits: light orange and hexagonal shape, fruit size traits: light red and square shape, agronomic traits: light green and rhombus shape, biochemical traits: blue and round shape. Acronyms for digital fruit size and shape traits are reported in Supplementary Table 1; acronyms for agronomic and metabolic traits are reported in Tables 1 and 2 , respectively. a ECW x SCM; ( b ) PER x ECW ddRAD-seq analysis High throughput sequencing of the ddRAD libraries constructed with Mbo I and Sph I endonucleases yielded a total of 241 Gb of data, corresponding to approximately 772 million demultiplexed reads with average coverage per individual ranged from 17× to 57×, with a mean of 32× (depth ~ 6%). On average, 90% of the reads were mapped onto the pepper reference genome with a range of 85–96% (Supplementary Table 3). Following alignment and SNP calling, 101,632 SNPs were preliminarily identified across the parents and F 2 individuals. For the ECW × SCM population applying a MAF of 0.05 and removing heterozygous markers in the parental lines and monomorphic markers across population resulted in 8,408 SNPs. The subsequent removal of 1,886 distorted and 1,346 redundant markers led to a final set of 5,176 markers utilized for the genetic map construction. Similarly, for the PER × ECW population, filtering with a MAF of 0.05 and eliminating heterozygous markers in the parental lines and monomorphic markers throughout the population, resulted in 8,009 markers. The further removal 1,779 distorted and 1,024 redundant markers has led to the final number of 5,206 markers used for the genetic map construction. Genetic maps construction The genetic maps in both F 2 populations were constructed using the filtered markers as described above. Details of distorted and segregated markers are provided in Supplementary Table 4. The majority of distorted markers exhibited a skew toward the male (pollen donor) parent, accounting for 81% and 88% of the skewed loci in the ECW × SCM and PER × ECW, respectively. In both cases, 12 linkage groups (LGs) corresponding to the 12 pepper chromosomes were identified. Following the initial iteration, 20 markers were eliminated from the ECW x SCM genetic map construction as they either mapped in different chromosomes (13) or were highly distant, thus distorting the map length (4) or exhibited inverted genetic versus physical positions. The final genetic map consisted of 5,156 markers spanning 2707.46 cM and defining 4,554 unique genetic bins (e.g. unique map position) (Table 4 ). For the PER × ECW population, 12 markers were removed the 58% of which were mapped on different chromosomes and the rest being highly distant or inverted along the chromosome. This map comprised a large number of SNPs (5,194) spanning 2589.57 cM and defining 4,641 genetic bins (Table 5 ). The ECW × SCM genetic map covered a slightly higher total genetic distance compared to the PER × ECW map, although this trend was not observed in all chromosomes. Indeed, five out of the twelve chromosomes were longer in the latter population. The chromosome length in the ECW × SCM map ranged from 176.86 cM (LG09) to 382.28 cM (LG03), while for the PER × ECW cross, lengths ranged from 97.5 cM (LG5) to 292.83 cM (LG01). Chromosomes 3 and 5 exhibited the greatest differences in length between the two populations. In both maps, the average inter-marker gap distance in all chromosomes did not exceed 0.7 cM, and the biggest gap was found on LG03, measuring 3.31 cM and 8.96 cM in ECW × SCM and PER × ECW, respectively. Table 4. Characteristics of the high-density genetic map for the ECW x SCM population Chr. Marker n ° Genetic distance (cM) Average gap (cM) Average inter-distance (cM) Max gap (cM) Bins Marker n ° Physical distance (Mb) Average inter-distance (Mbp) Average gap (Mbp) Max gap (Mbp) R 2 LG01 488 238.18 0.55 0.49 2.16 433 492 332.15 0.68 0.67 22.44 1.00 LG02 497 266.67 0.62 0.54 2.13 430 497 177.22 0.35 0.35 40.50 0.99 LG03 677 382.28 0.63 0.57 8.96 608 678 289.27 0.43 0.43 40.07 0.99 LG04 355 198.43 0.62 0.56 2.44 321 357 245.56 0.69 0.69 56.41 0.89 LG05 372 215.40 0.66 0.58 2.14 325 374 254.01 0.68 0.68 30.63 0.99 LG06 426 211.34 0.58 0.50 3.50 368 429 252.70 0.59 0.59 41.57 0.99 LG07 398 202.28 0.6 0.51 1.95 339 398 265.58 0.67 0.67 35.56 1.00 LG08 354 190.27 0.58 0.54 2.29 330 354 173.90 0.49 0.49 31.80 0.99 LG09 380 176.86 0.53 0.47 1.84 334 383 277.25 0.73 0.72 55.91 0.99 LG10 373 191.27 0.6 0.51 2.18 322 375 209.55 0.56 0.56 58.60 1.00 LG11 472 230.77 0.54 0.49 2.66 425 473 274.72 0.58 0.58 21.14 0.97 LG12 364 203.70 0.64 0.56 3.33 319 366 259.67 0.71 0.71 35.55 0.99 Open in a new tab R 2 Spearman based determination coefficient between genetic map and physical map Table 5. Characteristics of the high-density genetic map for the PER x ECW population Chr. Marker n ° Genetic distance (cM) Average inter-distance (cM) Average gap (cM) Max gap (cM) Bins Marker n ° Physical distance (Mbp) Average inter-distance (Mbp) Average gap (Mbp) Max gap (Mbp) R 2 LG01 506 292.83 0.58 0.64 2.98 462 509 332.15 0.65 0.65 26.26 0.90 LG02 530 261.89 0.49 0.55 1.80 477 530 177.22 0.33 0.33 30.24 0.99 LG03 632 288.27 0.46 0.51 3.31 566 633 289.23 0.46 0.46 27.44 1 LG04 274 137.59 0.50 0.55 2.00 253 276 245.70 0.89 0.89 42.13 0.99 LG05 245 97.05 0.40 0.45 1.57 217 246 252.80 1.03 1.02 82.90 0.99 LG06 440 243.13 0.55 0.64 2.96 383 441 252.70 0.57 0.57 25.15 0.99 LG07 447 195.14 0.44 0.49 1.71 399 447 266.22 0.60 0.6 68.87 0.99 LG08 365 150.14 0.41 0.47 1.67 323 365 173.90 0.48 0.48 103.91 0.99 LG09 475 252.51 0.53 0.61 2.33 414 476 277.02 0.58 0.58 29.03 0.88 LG10 432 211.97 0.49 0.55 2.69 385 432 209.17 0.48 0.48 7.37 0.96 LG11 490 256.99 0.52 0.6 2.04 431 493 274.39 0.56 0.56 22.02 0.93 LG12 358 202.06 0.56 0.61 2.25 331 358 259.58 0.73 0.72 38.64 0.99 Open in a new tab R 2 Spearman based determination coefficient between genetic map and physical map The physical maps were nearly identical, measuring approximately 3.000 Mbp and differing by only 1.51 Mbp, with discrepancies greater than 1 Mb only on chromosome 5. While the average genetic distance between markers was comparable in both populations (less than 0.53 cM and 0.58 Mbp in the genetic and physical maps, respectively), larger gaps were found considering the physical position. To assess the quality and accuracy of the genetic maps, the genomic positions of the bin markers were compared with the Capsicum annuum var. Zhangshugang reference genome. A high degree of collinearity was observed between the genetic maps and their respective physical positions, as evidenced by Spearman rank correlation coefficients of 0.86 and 0.80 for the ECW × SCM and PER × ECW map, respectively. Very few regions displayed inconsistences across the chromosomes (Fig. 5 ); this occurred specifically on chromosome 11 in the ECW × SCM map and on chromosomes 9, 10 and 11 in the PER × ECW map. Collinearity details for each chromosome are reported in Supplementary Fig. 3, where inverted regions are identified as linear segments of points deviating from the main collinearity curve. The two maps provided comprehensive coverage of pepper genome, thus meeting the requirements for high-resolution QTL mapping. Fig. 5. Open in a new tab Collinearity of genetic and physical maps. a curves representing the relationship and collinearity for each chromosome of the ECW x SCM population, the x-y axes indicate the physical and genetic position of each ddRAD-SNP marker, respectively. b comparison of physical (blue bars) and genetic (red bars) position for the ECW x SCM cross. The solid grey lines connect each marker of the two maps and for each chromosome are in scale positioned in Mbp (left) and cM (right). c and d report the collinearity and comparison between of physical (blue bars) and genetic (red bars) position for the PER x ECW cross A consensus map was then constructed using 1969 markers shared between the two populations. The resulting map spanned 2.309 cM (Fig. 6 ), with chromosomes 1 (275.89 cM) and 3 (247.37 cM) as the longest one, while chromosomes 4 (158.99 cM) and 8 (137.07 cM) as the smallest. The consensus genetic map exhibited a very high degree of correlation to the physical one with Spearman rank correlation coefficients of 0.87. This map was provided the framework to localize QTL clusters for agronomic, metabolic and fruit morphometric traits. Fig. 6. Open in a new tab Consensus map based on 1969 common markers between the ECW × SCM and PER × ECW populations. The integrated map was calculated with Joinmap 4.0 and the regression mapping algorithm. a whole genome collinearity map, curves representing the relationship and collinearity for each chromosome of the consensus map, the x-y axes indicate the physical and genetic position of each ddRAD-SNP marker, respectively; ( b ) details for each chromosome: dots represent the markers positioned along the two maps, the orange lines indicate the connection between the genetic maps, and the length indicates the markers that appear to have a different ranking on the respective chromosomes based on genetic (cM) and physical position (Mbp) positions, the green line represent the collinearity curve; ( c ) synteny map comparing physical (green bars) and genetic (orange bars) positions, the solid grey lines connect each marker of the two maps and for each chromosome are in scale positioned in Mbp (left) and cM (right) QTL mapping analysis QTL for agronomic traits A total of 25 significant QTLs were found for agronomic traits, six of which for fruit width and the colour coordinates a*, b*, chroma in the ECW × SCM population and 19 for the colour coordinates L* and b* as well as for fruit weight and fruit length in the PER × ECW population (Supplementary Table 5). These QTLs were distributed across all chromosomes a part chromosome 5, and 12 (Fig. 7 ). Two major QTLs on chromosome 6 were found for the color b* coordinate; qb6.1 spanned a region of 33.29–34.91 cM with a LOD of 29.5 and a PVE of 58.2%, whilst downstream, the qb6.2 was positioned in a narrow region of 44.12–44.16 cM exhibited a LOD of 12.7, accounting for the 20.1% of the phenotypic variation. On the same chromosome was positioned a main QTL for chroma, qChr6.1 , within the 155.39–164.12.39.12 cM interval and contributing with a 13% of PVE. Three additional major QTLs were identified for fruit weight: qFW1.1 with a 6.5 LOD and 13.30% PVE was positioned in a 3 cM at the distal end of the chromosome 1, qFW2.2 with LOD 12.3 was localized on the bottom of chromosome 2 in a 0.57 Mbp region within the 215.55–225.26.55.26 cM interval and exhibited a PVE of 17.6%, qFW3.1 displaying a LOD score of 10.3 was positioned in a 1 cM interval on the upper arm of chromosome 3 and accounted for the 14.7% of the phenotypic variability. Also, for the length of fruits three main QTLs mapping on chromosomes 3 and 4 were detected: qFL3.1 (LOD = 8.5) and qFL3.2 (LOD = 14.8) were located on the upper and pericentromeric regions of the chromosome, contributing 7.47% and 15.70% to the PVE, respectively; qFL4.1 , with a LOD of 6.6 and PVE of 5.6% mapped at the bottom of chromosome 4. All of these major QTLs were found in the PER × ECW cross, except for qCt6.1 . Furthermore, they were associated with an increase in trait values by exhibiting positive additive effects, except for qFL3.2 , which displayed a negative one. Fig. 7. Open in a new tab QTL mapping analysis. Circos plot diagram showing the position of 200 QTLs detected in this study. The dashed red concentric lines indicate the threshold LOD = 3 with the relative LOD scale on the y-axis. For each chromosome is showed the integrated map for the two populations with relative markers density according to the coloured legend. The legend scale indicates the number of SNPs within 1 Mbp window size. QTLs are represented by coloured dots. a QTLs for metabolic traits in the ECW x SCM population, ( b ) QTLs for fruit shape traits in the ECW x SCM population, ( c ) QTLs for agronomic traits in the ECW x SCM population, ( d ) QTLs for metabolic traits in the PER x ECW population, ( e ) QTLs for fruit size traits in the PER x ECW population, ( f ) QTLs for fruit shape traits in the PER x ECW population, ( g ) QTLs for agronomic traits in the PER x ECW population The dashed grey lines delimit the position of each QTL highlighting pleiotropic and shared QTLs between populations QTL for fruit size and shape For imaging-based fruit size and shape parameters, QTLs were identified for all traits except for Distal Indentation Area. In total, 127 QTLs were detected for 37 parameters in the range of 1 for the distal angle micro, obovoid, proximal eccentricity, and shoulder height attributes to 9 for distal fruit blockiness. QTLs were classified based on the category of the acquired parameters. Following this criterion, 31 QTLs were detected for seven fruit size-related traits and 96 QTLs for 30 fruit shape-related traits (Supplementary Table 6). Fruit size QTLs were identified only in the PER × ECW population within different chromosomal clusters (Fig. 7 ). Single clusters associated with increased fruit size were identified in the distal region of chromosomes 1, 2 and 9, while on chromosome 6, three QTLs related to fruit height parameters were in the apical part. Among these, the cluster on chromosome 2 included 5 QTLs for a same number of fruit size parameters, all exhibiting a major effect with LOD values between 6.9 and 8.6 and PVE% in the range 14.6–24.4. The highest number of QTLs was found on chromosome 3. Two main clusters were detected; the first included 3 QTLs for fruit height in the interval 48.46–49.59 cM, the second included five QTLs and two subclusters in a large region spanning 130–150 cM, all with a major effect on the increase of fruit size by exhibiting a PVE from 18.3 to 32.6%. For fruit shape, QTLs were detected in both ECW × SCM and PER × ECW populations and all contributed to an increase of the assayed traits. The highest number of QTLs were found on chromosomes 1 and 3. In the first case, most QTLs were in the intervals 170–196 cM and 242–256 cM all with moderate effects not exceeding 10% of PVE. QTLs with major effects were instead identified on chromosome 3: for distal fruit blockiness, fs.dfb.3.1 and fs.dfb.3.2 (LOD 16.5 and 20.7, respectively) accounted for 13.1% and 16.6% of PVE, respectively; six QTLs for fruit shape index parameters located in the 130.31–150.16.31.16 cM region, displayed LOD in the range 23.1–27.5 with a PVE greater than 48%. Two additional QTLs responsible for the variation of ellipsoid ( fs.e.3.1 , LOD 19.7) and pericarp thickness ( fs.pt.3.1 , LOD 18.7) both contributed for the 36% of phenotypic variation. Three other clusters of approximately 1.5 cM each were found on chromosome 4 in the regions 41.29–42.66 cM and 50.01–51.48 cM and on chromosome 9 in the interval 213.45–214.85.45.85 cM. All of these, alongside the QTL identified on the remaining chromosomes, showed a minor or moderate effect with a PVE on average not exceeding 15% and LOD score in most cases lower than 5. QTL for metabolites Forty-eight QTLs were identified for the metabolites studied, 28 of which were mapped in the ECW × SCM population while the remaining 20 in the PER × ECW cross (Fig. 7 , Supplementary Table 7). Apart from quinate, where QTLs were only found in the ECW × SCM cross, both populations contributed to the detection of QTLs. A total of 22 QTLs were detected for organic acids, primarily for citrate (8) and malate (5). These two compounds exhibited the strongest peaks on chromosome 1: qcit1.1 (LOD of 9.1) was positioned at the distal end of the chromosome in the interval 222.91–225.60 cM and associated with a consistent decrease in the trait, explaining 18.7% of the PVE; at 70 cM distance from this QTL, another, qmal1.1 (LOD 5.8), determined a robust increase of trait performance displaying a PVE% of 17.9% and additive effect of 530.8. Five out of the 8 QTLs identified for citrate, contributed to the reduction of the content of this compound with an additive effect ranging from. −119.72 ( qcit1.1) to −20.9 ( qcit2.1 ). On chromosomes 10 and 12 were located the remaining three with moderate effect contributing to the increase of citrate with a PVE ranging from 6.2% to 11.6%. As for malate, beyond the major QTL above described four minor ones (PVE 1.5–4.5%) were detected and were responsible for both increase ( qmal5.1 and qmal8.1 ) and decrease ( qmal2.1 and qmal7.1 ) of the content of this compound. Four QTLs were identified for ascorbate, the strongest ones located at the bottom of chromosome 1 and the top of chromosome 6 explaining 12.7 and 17.8% of the PVE, respectively and both contributing to the decrease of the trait. Between the two QTLs for quinate, the main one, qqui4.1 , was positioned on chromosome 4 in the interval 106.11–109.80 cM and contributed with increasing quinate content by explaining 14.41% of PVE. For succinic acid, 3 QTLs with moderate effect explaining similar PVE (~ 11%) were detected, and only one ( qsuc9.1 ) contributed to the increase in succinate levels. For the three sugar compounds analyzed, a total of 21 QTLs were identified, including eight for fructose, six for glucose, and seven for sucrose, mostly with moderate or small effects. Within fructose, the two major QTLs were found on chromosome 11 at 54 cM apart: qfru11.1 showed a LOD of 5.1 and PVE of 12.9%, while qfru11.2 displayed a LOD of 4.3 and PVE of 10.94%. Both loci were associated with a reduction in fructose content. Another cluster on chromosome 3 was found in the broad interval (205–382 cM) spanning over 20 Mbp and included three QTLs contributing to both increase and decrease of the traits with a PVE in the range 3.5–8.3%. For glucose, qglu2.1 was the main QTL (LOD 6.6) located in the 12.51–13.18 cM interval region on chromosome 2, which increased the trait with a PVE of 16.7%. The other QTLs, mapping to the pericentric regions of chromosomes 1, 2, 7 and 11 exhibited moderate effects with PVEs ranging from 5.6% to 11.5% and showing both positive and negative additive effects. Sucrose variation was supported by two major QTLs located on the upper arms of chromosome 8, ( qsac8.1 , LOD 6.9), and 10 ( qsac10.1 , LOD 4.0). The former decreased sucrose content with a PVE of 6.6%, while the latter showed the opposite trend with a PVE of 16.4%. Other QTLs with low or moderate effects (LOD 3.0–3.7.0.7) with PVE in the range of 1.5–4.8% were found on chromosomes 1, 5, and 10. For β-carotene, five QTLs were identified as being associated with a decrease in the trait. Four of these were located on chromosome 6 in different interval regions that ranging in size from 0.8 cM to 10 cM, while the fifth was located on chromosome 4 in a 5 Mbp region corresponding to a 50 cM interval. The strongest of them, qcar6.2 (LOD = 3.7), showed a PVE of 18.8%. QTL clusters and pleiotropic regions underlying agronomic, qualitative and fruit size and shape traits A total of 113 QTLs underlying either metabolic, agronomic and fruit size and shape traits converged into 23 clusters distributed across different chromosomes (Fig. 8 , Supplementary Table 8). These QTLs were positioned on the integrated map comprising 1969 common markers between the ECW x SCM and PER x ECW populations. QTLs that were present in both populations were included in nine clusters. This co-localization may reflect either pleiotropy or tight linkage, a common feature in genomic regions with high QTL density. Among them, a QTL cluster for agronomic and fruit size and shape traits was established at the bottom of chromosome 2 with a pleiotropy at the peak area at 163 Mbp; citrate and fruit shape traits were found to colocalize at 170 cM on the same chromosome. Another cluster of fructose, citrate and fruit size traits was detected in the broad interval at the bottom of chromosome 3. On chromosome 6, QTLs for eccentricity colocalized the color coordinate b* and beta-carotene at 18.7–29.6 cM. The latter two traits also colocalized with fruit weight at 44–48 cM on the same chromosome. On chromosome 11, another cluster was responsible for the increase of fruit weight, citric acid and succinate, similarly on chromosome 12 citric acid colocalized with maximum width in the interval 17.5–25.6 cM. Pleiotropic QTLs were found for fruit weight and heigh-mid width and 49 cM on chromosome 3 as well as for fruit length and different digital fruit shape attributes at 120 cM. Fig. 8. Open in a new tab Quantitative trait loci pleiotropy map displaying the location of QTL clusters controlling different trait categories. The horizontal bars indicate the range (cM) based on distances calculated on the integrated map that including 1969 SNP markers common to the two populations. The integrated map was calculated with Joinmap 4.0 and the regression mapping algorithm. The 23 clusters are defined by the overlapping vertical bars delimiting the interval. The name of the QTL belonging to the cluster is indicated next to each vertical bar. Details for each QTL are reported in Supplementary Tables 5, 6, and 7. The different category of QTLs are indicated following the color coding displayed in the legend Considering the QTL clusters detected in a single population, several chromosomal regions showed pleiotropic QTLs for metabolites and for traits associated to fruit weight and morphometry. Malic acid co-mapped with fruit shape traits at 147 cM on chromosome 1, with perimeter in the interval 38–41 cM on chromosome 5, with glucose, eccentricity, v. asymmetry and pericarp thickness at the bottom of chromosome 7, and with sucrose, fruit length and v. asymmetry at the top of chromosome 8 at 5.1 cM. Ascorbic acid colocalized with fruit weight and fruit shape on chromosome 1 in the interval 240–243 cM and with β-carotene on chromosome 6 in the interval 98–100 cM. As for the genetic effect, the cluster at the distal end of the chromosome 3 stands out as the most impactful, containing 21 QTLs exhibiting a consistent additive mode of action and synchronized directional effects: alleles for increased fruit size were consistently associated with specific fruit shape determinants supporting a highly coordinated developmental control at this locus. In contrast, metabolic clusters showed more heterogeneous inheritance patterns. For instance, the hotspot on chromosome 3, including QTLs for sugar content and citric acid, exhibited primarily dominance or overdominance effects. Interestingly, some clusters revealed cross-category co-localizations, such as those one on chromosome 1, where increased fruit width and weight were linked to a decrease in Vitamin C content under an overdominant model, suggesting potential metabolic trade-offs during fruit expansion. Candidate genes In total, 18 candidate genes were found to underlie 28 out of the 200 discovered QTLs (Table 6 ). Most genes were found for fruit size and shape traits, for which, several binding proteins were found associated. Notably, the ABC binding transporter Caz03g05770 was responsible for the pleiotropic cluster on chromosome 3. Furthermore, a putative acyl-activating enzyme Caz01g30730 , the homeobox associated leucine zipper protein HAT5 Caz02g31220 and the RNA-binding protein FUS gene Caz04g0717 0 — with protein, DNA and RNA binding functions, respectively — were associated with QTLs on chromosome 1, 2 and 4. Two transcription factors Caz05g00420 (encoding for a RNA polymerase II transcription subunit) and Caz09g00450 (encoding for a translation elongation factor), underlie the QTLs for fruit shape on chromosomes 5 and 9, respectively. An adenylosuccinate synthetase Caz02g23760 was linked to the QTL for fruit weight qFW2.2 while on chromosome 4, the CPA3 gene ( Caz04g00830 ) possessing metallocarboxypeptidase activity and a detoxification protein ( Caz04g20470 ) with antiporter activity — both reported in C. annuum — underlie qL4.1 and qFL4.1 , respectively. Regarding metabolic traits, fructose and malic acid levels were linked to three candidate genes: Caz02g08940 , which encoded for a proton transporter ATPase , for qfru2.1 , while Caz03g38670 , encoding for a glutathione reductase , and Caz12g02400 , which encoded for a kinesin-like protein , were associated with qcit3.2 and qcit12. 1, respectively. Table 6. Candidate genes identified for QTLs underlying agronomic, fruit morphology and qualitative traits across two intraspecific F 2 populations QTL Gene Function Gene Ontology Category Agronomic qFW2.2 Caz02g23760 Adenylosuccinate synthetase, chloroplastic adenylosuccinate synthase activity molecular function qL4.1 Caz04g00830 mast cell carboxypeptidase A isoform X1 ( Capsicum annuum ) CPA3 metallocarboxypeptidase activity molecular function qFL4.1 Caz04g20470 Protein DETOXIFICATION 33 isoform X3 ( Capsicum annuum ) antiporter activity molecular function Fruit shape fs.dfb.1.3 Caz01g30730 Putative acyl-activating enzyme 19 protein binding molecular function fs.asov.2.2 Caz02g31220 Homeobox associated leucine zipper protein HAT5 DNA-binding transcription factor activity, RNA polymerase II-specific molecular function fs.r.3.1 Caz03g23960 Carboxyl-terminal-processing peptidase 3, chloroplastic protein binding molecular function fs.c.3.1 Caz03g05770 ABC transporter C family member 13 ATP binding molecular function fs.fsc.3.1 Caz03g05770 ABC transporter C family member 13 ATP binding molecular function fs.fse.3.1 Caz03g05770 ABC transporter C family member 13 ATP binding molecular function fs.fsi.3.1 Caz03g05770 ABC transporter C family member 13 ATP binding molecular function fs.asov.3.1 Caz03g05770 ABC transporter C family member 13 ATP binding molecular function fs.ld.3.1 Caz03g05770 ABC transporter C family member 13 ATP binding molecular function fs.pma.3.1 Caz03g05770 ABC transporter C family member 13 ATP binding molecular function fs.asv.3.1 Caz03g05770 ABC transporter C family member 13 ATP binding molecular function fs.dfb.3.1 Caz03g17860 PH%2 C RCC1 and FYVE domains-containing protein 1-like ( Solanum pennellii ) fs.dma.4.1 Caz04g05870 Syntaxin-6_N domain-containing protein membrane cellular component fs.ob.4.1 Caz04g07170 RNA-binding protein FUS-like ( Capsicum annuum ) molecular function fs.pma.4.1 Caz04g07170 RNA-binding protein FUS-like ( Capsicum annuum ) molecular function fs.e.5.1 Caz05g00420 Mediator of RNA polymerase II transcription subunit 13 transcription coregulator activity molecular function fs.e.5.2 Caz05g17940 Folylpolyglutamate synthase tetrahydrofolylpolyglutamate synthase activity molecular function fs.dfb.9.1 Caz09g00450 Elongation factor Ts translation elongation factor activity molecular function Fruit size fz.p.3.1 Caz03g05170 carbonyl reductase [NADPH] 1 oxidoreductase activity, acting on the CH-OH group of donors, NAD or NADP as acceptor molecular function fz.mw.3.1 Caz03g05770 ABC transporter C family member 13 ATP binding molecular function fz.mw.4.1 Caz04g07170 RNA-binding protein FUS-like ( Capsicum annuum ) molecular function fz.a.4.1 Caz04g24290 UDP-N-acetylglucosamine transferase subunit ALG14 homolog dolichol-linked oligosaccharide biosynthetic process biological process Metabolites qfru2.1 Caz02g08940 mRna(V-type proton ATPase subunit F) proton-transporting V-type ATPase, V1 domain cellular component qcit3.2 Caz03g38670 Glutathione reductase glutathione-disulfide reductase (NADPH) activity molecular function qcit12.1 Caz12g02400 Kinesin-like protein microtubule motor activity molecular function Open in a new tab Discussion Genetic maps of pepper as frame for QTL discovery Pepper is a significant source of nutritional compounds beneficial for human health which, together with the weight, colour, size and shape of the fruit, constitute the key factors determining yield and quality and main targets for breeding. This work aims to gain new insights into the genetic basis of pepper quality traits by examining a wide range of traits and their complex interaction to identify previously undiscovered QTLs and pleiotropic regions that might benefit breeding programs. Establishment of mapping populations and development of linkage maps represent the foundation to investigate the genetic architecture of these complex traits. In this study, two F 2 populations were created using highly divergent founders selected based on a global survey of the existing phenotypic and genetic diversity in pepper [ 3 ]. The resulting phenotypic variation observed in these segregating populations proved to be reliable for QTL mapping. F 2 ’s are powerful to dissect QTLs with additive and dominant effects although drawbacks may occur in mapping resolution [ 17 ]. In this context, high-throughput genotyping methods enhance the enrichment of maps with genomic markers, enabling increased accuracy in QTL identification [ 51 ]. Furthermore, the parallel analysis of interrelated biparental families provides a greater allelic variation sampling and increases the power to identify common QTLs [ 51 ]. To this end, a common parental line was used for the construction of the two populations and ddRAD-seq employed as high throughput genotyping strategy to implement linkage maps construction and QTL identification. This technology effectively reduced genome complexity and enabled high-resolution marker discovery on a large scale [ 52 ]. The application of a double endonuclease system and fragment size selection, allow a reliable detection of loci across individuals, thus reducing missing calls and genotyping errors. Furthermore, paired-end sequencing allows to accurately map reads onto a reference genome [ 33 ]. Compared with previous maps established in the Capsicum genus through SNPs markers [ 28 , 53 – 56 ], the maps here reported constituted the first example that used ddRAD-seq in pepper intraspecific F 2 ’s and provide both greater length and a smaller number of gaps in terms of number and size. Additionally, the current study exhibits a higher bin density with respect to earlier findings. The two genetic maps facilitated a more precise definition of chromosome lengths relative to physical positions. In agreement with previous investigations, chromosome 3 was the largest, holding the highest number of bins and distorted markers [ 54 – 57 ]. Markers in pericentric regions showed greater segregation distortion. This phenomenon, which is frequent in biparental populations with a high genetic differentiation, results from deviations of allelic segregation ratios from the expected Mendelian proportions. It is driven by mechanisms such as the non-random allele segregation related to the selection against certain alleles during gametogenesis (gamete competition) or after fertilization (zygotic selection) [ 58 ]. The percentage of distorted markers typically increases in interspecific crosses or in crosses between highly distant founders. The observed segregation distortion of this study is in line with previous findings in pepper [ 54 , 55 , 59 , 60 ], which showed a trend varying from 20 to 40%, thus confirming how this would not affect the estimation of the recombination rate and genetic distance for QTL mapping. Higher numbers of markers were skewed toward the male parent in both populations. This was consistent with previous studies in pepper including Barchi et al. [ 54 ], Cheng et al. [ 60 ], Zhu et al. [ 56 ] but not with Zhang et al. [ 54 ]. Nuclear-cytoplasmic interactions, incompatibility mechanisms, and pollen competition [ 61 , 62 ] may account for this bias, as observed in other crops where uneven parental contribution occurs even at the chromosomal level [ 63 , 64 ]. According to earlier research, using a common genotype as the male parent in a controlled cross results in a larger segregation distortion than using it as the female parent [ 61 ]. This is consistent with this study’s findings, which showed that the population with ECW as a male parent had greater biased markers. However, Bodénès and colleagues [ 61 ], also considered interspecific crosses, which could have contributed to the reported distortion. Nevertheless, results of the current work supported a distortion unbalanced toward one parent in bi-parental crosses. Further investigations are required to elucidate the biological processes driving uneven segregation in pepper breeding. The average Spearman correlation coefficients ​​observed above 0.97 in both populations were higher than those reported in previous findings [ 55 , 56 ], thus highlighting a high collinearity that was found between the genetic and physical positions on the pepper genome. This high collinearity combined with dense marker maps ensures robust and accurate identification of QTLs and facilitates the reliable discovery of candidate genes. QTL analysis and candidate genes detection In this study, novel agro-qualitative QTLs were detected evaluating 54 phenotypic traits related to fruit’s basic agro-morphological parameters and nutritional content. The analysis highlighted QTLs with small and large effects as well as QTL clustering and presence of pleiotropic loci. Most QTLs revealed a complex genetic architecture characterized by a clear divergence between morphological and metabolic traits. While agronomic and morphological traits were primarily governed by additive and partial dominance effects, several quality-related traits exhibited dominance or overdominance. This distinction suggests that while structural development follows a dosage-dependent allelic model, bioactive compounds are driven by more complex allelic interactions. These results agree with previous studies highlighting that a higher percentage of metabolic QTLs (mQTLs) often show complex, non-additive inheritance, exhibiting a high proportion of dominance effects [ 65 – 67 ]. Across the 200 QTLs identified in the two populations, a set of 20 candidate genes localized in the peak region of 28 QTLs were identified. For remaining QTLs, candidate genes were in most instances at less than 100 kilobases apart. The molecular function was examined considering pepper annotation and comparing this with functions in other organisms. Despite their distribution throughout the genome, the analysis revealed few hot spots. Most QTLs contributed to the variation of fruit size and shape traits; this trend was expected given the high morphological diversity of the parental lines. Chromosome 3 holds the highest number of QTLs exhibiting both high PVE and LOD scores as well as candidate genes for fruit shape. This highlights the presence of a genomic hotspot of high-impact QTLs, suggesting that these traits are governed by major genes rather than a complex polygenic architecture. In line with the dosage-dependent allelic model, these QTLs exhibited a purely additive mode of action, confirming that fruit architecture scales linearly with allele dosage. As for the high proportion of phenotypic variance explained, any potential overestimation (the Beavis effect) can be regarded as negligible for these high-impact loci. As demonstrated by Xie et al. [ 68 ], the Beavis effect diminishes significantly as the magnitude of the test statistic increases. Given the exceptionally high LOD scores observed and the absence of missing data—which ensured a stable effective sample size—these QTLs can be considered robust biological signals rather than statistical artifacts. Among candidate genes, a cluster of ABC transporters involved in the transfer of key metabolites underlying growth and developmental processes in plants was identified [ 69 ]. These genes are generally highly expressed during early fruit development due to their role in auxin transport, and their downregulation can lead to fruit malformation [ 70 , 71 ]; additionally, their role in the evolution of seed size has been also reported [ 72 ]. A high number of ABCs transporters has been previously reported on the chromosome 3 of Capsicum annuum [ 73 ], a region known to harbour the largest number of the fruit shape-related genes [ 53 , 74 , 75 ]. Furthermore, an RCC1 ( regulator of chromosome condensation 1 ) and FYVE like domains were found associated with distal fruit blockiness; these gene are mostly involved in plant development and response to environmental stresses [ 76 ]. HAT5 was also found associated to fruit asymmetry at the bottom of chromosome 2. This gene, part of the homeobox-associated leucine zipper (HD-Zip) protein family, is involved in several plant growth and development mechanisms including apical embryonic upper-tier cells and embryonic cotyledons development [ 77 ]. Furthermore, HAT5 enhances the resistance to salt stress and drought by controlling proline metabolism and reducing reactive oxygen species generation [ 78 ]. Chromosomes 2 and 3 are those where a high number of QTLs have been found in pepper biparental populations [ 27 , 74 , 79 , 80 ]. Several organ size and shape gene homologs have been reported at the distal ends of both chromosomes [ 57 ]. Specifically, HAT5 is positioned at 2Mbp from Longifolia 1 [ 74 ] and in the same interval of CLAVATA3 , a gene underlying the fas mutation [ 81 ]. By regulating cell organization and differentiation processes during early and late fruit development, these genes contribute to the diversity in fruit morphology [ 82 ]. Furthermore, the candidates identified on chromosome 3 overlapped with the fsi locus responsible of a non-synonymous mutation in the TRM (TONNEAU 1 Recruiting Motif) protein Capana03g002426 which control cell division patterns regulating of fruit shape [ 83 ]. Another candidate on chromosome 1, Caz01g30730 encoding for a putative acyl-activating enzyme was reported to be involved in the activation of carboxylate, key intermediates in many metabolic pathways for central and specialized metabolites [ 84 ]. The RNA-binding protein FUS gene Caz04g07170 has been also found involved in both fruit size and shape traits on chromosome 4; although there is no specific function related to these traits, this gene overlap with previous identified QTLs for locule number and fruit shape [ 24 , 85 ]. For all these genes, the direct link between with pepper fruit shape and size has not yet been established however, the robust associations in agreement with previous findings suggest their involvement. The function of these gene should be further explored, and their putative role in stress resistance could be exploited for the development of new stress-resistant crop varieties. The QTL qFW2.2 identified in this study co-localized with fw2.1 a QTL with large effect reported to increase fruit weight and explaining 62% of the trait variation [ 24 ]. qFW2.2 was positioned near the Ovate gene, which controls fruit shape, and mapped within the same interval as several fruit size and shape QTLs [ 27 , 53 ]. This demonstrates that this QTL is potentially involved in fruit weight gain. The underlying candidate gene, encoding an adenylosuccinate synthetase , plays a critical role in the accumulation of nutrients maize seeds, thus regulating their development [ 86 ]. Although direct evidence in pepper is currently lacking, the genomic position of qFW2.2 and its association with fruit shape traits support a potential role in the mechanisms that regulating berry weight variation. Three candidate genes were found responsible for the increase of fructose and citric acid. On chromosome 2, a proton-transporting V-type ATPase , essential for maintaining proton gradients across membranes and involved in multiple physiological functions [ 87 ], underlie qfru2.1 . Although the role of these genes is mostly related to maintaining cellular homeostasis—with possible implications for adaptation to environmental stress— their involvement in citrate accumulation has been documented in citrus [ 88 ]. This suggests a potential role in the accumulation of other primary metabolites in other horticultural crops. Instead, the build-up of citrate in fruits, was caused by a glutation reductase and a kinesin-like protein which underlie qcit3.2 and qcit12.1 , respectively. The first candidate has been primarily implicated in bacterial defence against citric acid rather than fruit accumulation [ 89 ], while the second is active in moving cellular components along microtubules. Therefore, the specific function of these candidates in regulating citric acid accumulation remains to be elucidated. Additional QTLs detected in the present study colocalized with those detected in a previous investigation aimed at decoding the architecture of the primary metabolome by combining a GWAS panel with a population of backcross inbred lines [ 32 ]. Specifically, qsac1.2 , qfru11.1 , and qfru11.2 , increasing the content of sucrose and fructose, mapped to the same regions on chromosomes 1 and 11 as reported by Von Steimker et al. [ 32 ]. Furthermore, qmal2.1 and qglu7.1 , which increased the contents of malic acid and glucose, were positioned 2 and 9 Mbp apart from the QTLs identified in the earlier work. These results suggest the existence of conserved regions governing nutritional quality in Capsicum , which could be exploited for further functional studies and precision breeding. While the identified loci align with known biological pathways, the reported genes should be considered as positional candidates. Further functional validation, such as gene expression profiling or CRISPR/Cas9-mediated genome editing, remains necessary to definitively confirm their causal roles. Co-located QTL clusters for different traits Several QTLs governing distinct traits were found to overlap in different genome regions. Colocalization is a common phenomenon due to pleiotropic effects of genes involved in various biological processes and or linked genes underlying QTLs [ 90 ]. Breeding programs can benefit from these QTL clusters with the possibility to simultaneously improve multiple traits. In the majority of these clusters, the additive effects exhibited consistent directions across co-localizing traits, particularly for fruit morphology. The colocalization mainly reflected the observed correlation between traits. Indeed, most pleiotropic regions involving positively correlated traits such as fruit weight, size and shape traits displayed similar additive effects. These findings suggest that these hotspots may reflect true pleiotropy, where a single regulatory gene coordinates multiple developmental processes. Furthermore, the identification of multi-trait clusters provides a deeper understanding of the genetic architecture of fruit agro-qualitative traits. The remarkable consistency observed on chromosome 3, where dozens of morphological QTLs share the same additive direction and high LOD scores, strongly supports the hypothesis of true pleiotropy. Such a pattern is typical of master regulatory genes that control fruit organogenesis through a dosage-dependent mechanism, where each allelic substitution contributes linearly to the final development of the fruit. For metabolites, colocalization with morpho-agronomic traits led to opposite effects, thus suggesting how these QTLs are not always potential to increase both the nutritional content and the size of the fruits. Such cases, where metabolic and morphological traits diverge in their additive directions, might instead be indicative of tight linkage between distinct causal genes rather than pleiotropy. Furthermore, the prevalence of overdominance and dominance in these regions, particularly for bioactive compounds like malic acid and AsA, suggests that metabolic traits are governed by intricate enzymatic or signaling interactions rather than simple structural scaling. However, clusters of interest have been found; for instance, the base of chromosome 3 was associated with an increase in sugar fruit content alongside fruit size while decreasing the citric acid content, making it a key region for determining the organoleptic and commercial characteristics of the berry. Similar favorable trends were found for the cluster at the top of chromosome 5, which increased malic acid content alongside fruit perimeter, and for clusters on chromosomes 9 and 12, which increased sucrose and citric acid content together with morphological parameters. Nevertheless, QTLs displaying opposite effects shouldn’t be considered adversely as they enable the selection of chili pepper varieties with specific nutritional profiles tailored to diverse market destinations (e.g. preserves, powder, packaging). The implementation of genome wide SNP data allowed the precise detection of these QTL clusters enhancing selection accuracy. While it is not possible to definitively rule out tight linkage between independent genes—especially in clusters with contrasting modes of action—the physiological correlation between fruit expansion and the regulation of specific metabolites points toward pleiotropic nodes that coordinate the resource allocation between fruit structural development and the synthesis of bioactive compounds. Further studies investigating the molecular mechanisms underlying the regulation of fruit size and metabolic traits will provide better knowledge of their relationships and effects. Conclusion The development of genetic maps based on two pepper segregating populations allowed the detection of novel QTLs underlying fruit quality, which may affect consumer and market acceptability as well as preference. Given the limited efforts made in Capsicum to integrate QTL analysis for fruit agronomic, morphological, and fruit metabolite features, the study seeks to close a gap when compared to other vegetable crops. Results confirm the complex nature of quality-related traits in pepper by discovering 200 novel QTLs and highlighting 23 hotspots that confirm previous findings while opening new ones. Findings suggests that fruit architecture in these populations is governed by master regulatory genes that can be precisely targeted for crop improvement. Furthermore, the identification of pleiotropic clusters coordinating fruit expansion with sugar and organic acid accumulation provides a roadmap for balancing yield with nutritional quality. The high-density genotyping analysis and the construction of genetic maps based on 5,000 quality markers for each F 2 population, allowed to refine the genomic regions of QTLs, which may help selecting attributes of commercial interest. These discoveries open up the opportunity by giving a thorough understanding of the genetic factors underlying fruit quality properties and offering robust targets for molecular breeding and genomic selection toward precision crop improvement in pepper. Supplementary Information Supplementary Material 1. (614.3KB, pdf) Supplementary Material 2. (139.1KB, xlsx) Acknowledgements The authors wish to thank Prof. Lorenzo Barchi and Dr. Luciana Gaccione for providing the revised annotation file of C. annuum cv. Zhangshugang. Abbreviations QTL Quantitative trait locus ddRAD-seq Double digest restriction-site associated DNA ECW Early California Wonder SCM Criollo de Morelos cv. CM334 PER Perennial MAF Minimum allele frequency ICIM Inclusive composite interval mapping LOD Likelihood of odds Authors’ contributions P.T. conceived the work and planned experiments; P.T., A.C. and R.M. managed greenhouse trials; A.C and P.T. performed genotyping experiment; A.C. performed agronomic and digital fruit size and shape acquisitions; A.C., A.D.A and G.F. performed metabolic analysis. Data were analyzed jointly by P.T. and A.C. All authors reviewed and edited the manuscript and approved the final version. Funding This study was supported was supported by the RGV-ORFLORA project funded by the Ministry of Agriculture, Food Sovereignty and Forests of Italy and by the Agritech National Research Center that received funding from the European Union Next-Generation EU (PIANO NAZIONALE DI RIPRESA E RESILIENZA (PNRR)–MISSIONE 4 COMPONENTE 2, INVESTIMENTO 1.4—D.D. 1032 17/06/2022, CN00000022). In particular, this study covers activities comprised in the Spoke 1, Task 1.2.1 Linking phenotype and genotype: discovery of loci/genes/alleles for traits of interest. Data availability The data have been uploaded to the National Center for Biotechnology Information (NCBI) with the BioProject accession PRJNA1403783 ([https://www.ncbi.nlm.nih.gov/bioproject/PRJNA1403783](https:/www.ncbi.nlm.nih.gov/bioproject/PRJNA1403783)) with individual accessions listed in Supplementary Table 3. The VCF file containing raw SNP data of the two mapping populations is available in Zenodo [https://zenodo.org/records/18133695](https:/zenodo.org/records/18133695) (doi: [10.5281/zenodo.18133695](https:/doi.org/10.5281/zenodo.18133695)). All the other data supporting our findings are available within the manuscript and the supplementary data. Declarations Ethics approval and consent to participate Not applicable: this article does not contain clinical trial and investigations involving human participants or animals performed by the authors. Experimental research and field studies on pepper plants comply with relevant institutional, national, and international guidelines and legislation. Consent for publication Not applicable. Competing interests The authors declare no competing interests. Footnotes Publisher’s Note Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations. References 1. Barboza GE, García CC, de Bem Bianchetti L, Romero MV, Scaldaferro M. Monograph of wild and cultivated chili peppers ( Capsicum L, Solanaceae). PhytoKeys. 2022;200:1. [ DOI ] [ PMC free article ] [ PubMed ] [ Google Scholar ] 2. Tripodi P, Kumar S. The Capsicum crop: an introduction. In: Ramchiary N, Kole C, editors. The Capsicum genome. Cham: Springer; 2019. [ Google Scholar ] 3. Tripodi P, Rabanus-Wallace MT, Barchi L, Kale S, Esposito S, Acquadro A, et al. Global range expansion history of pepper ( Capsicum spp.) revealed by over 10,000 genebank accessions. Proc Natl Acad Sci U S A. 2021;118(34):e2104315118. [ DOI ] [ PMC free article ] [ PubMed ] [ Google Scholar ] 4. FAOSTAT. 2025. https://www.fao.org/faostat/en/#data Accessed 07 January 2026. 5. Devi J, Sagar V, Kaswan V, Ranjan JK, Kumar R, Mishra GP, et al. Advances in breeding strategies of bell pepper ( Capsicum annuum L. var. grossum Sendt) in advances in plant breeding strategies: vegetable crops cham. Springer Int Publishing. 2021;9:3–58. [ Google Scholar ] 6. Parisi M, Alioto D, Tripodi P. Overview of biotic stresses in pepper ( Capsicum spp.): sources of genetic resistance, molecular breeding and genomics. Int J Mol Sci. 2020;21(7):2587. [ DOI ] [ PMC free article ] [ PubMed ] [ Google Scholar ] 7. Kleih AK, Sparke K. Visual marketing: the importance and consumer recognition of fruit brands in supermarket fruit displays. Food Qual Prefer. 2021;93:104263. [ Google Scholar ] 8. Schreinemachers P, Simmons EB, Wopereis MC. Tapping the economic and nutritional power of vegetables. Glob Food Sec. 2018;16:36–45. [ Google Scholar ] 9. Tripodi P, Ficcadenti N, Rotino GL, et al. Genotypic and environmental effects on the agronomic, health-related compounds and antioxidant properties of chilli peppers for diverse market destinations. J Sci Food Agric. 2019;99(10):4550–60. [ DOI ] [ PubMed ] [ Google Scholar ] 10. Kondo F, Kumanomido Y, D’Andrea M, Palombo V, Ahmed N, Futatsuyama S, et al. Prediction of fruit shapes in F1 progenies of chili peppers ( Capsicum annuum ) based on parental image data using elliptic Fourier analysis. Comput Electron Agric. 2025;236:110422. [ Google Scholar ] 11. Wang L, Zhong Y, Liu J, Ma R, Miao Y, Chen W, Zheng J, Pang X, Wan. Pigment biosynthesis and molecular genetics of fruit color in pepper. Plants 2023; 11:2156. [ DOI ] [ PMC free article ] [ PubMed ] [ Google Scholar ] 12. Olatunji TL, Afolayan AJ. The suitability of chili pepper ( Capsicum annuum L.) for alleviating human micronutrient dietary deficiencies: a review. Food Sci Nutr. 2018;6:2239–51. [ DOI ] [ PMC free article ] [ PubMed ] [ Google Scholar ] 13. Sim M, Hong S, Jung S, Kim JS, Goo YT, Chun WY, et al. Vitamin C supplementation promotes mental vitality in healthy young adults: results from a cross-sectional analysis and a randomized, double-blind, placebo-controlled trial. Eur J Nutr. 2022;61:447–59. [ DOI ] [ PMC free article ] [ PubMed ] [ Google Scholar ] 14. Kashif M, Yao H, Schmidt S, Chen X, Truong M, Tüksammel E, et al. ROS-lowering doses of vitamins C and A accelerate malignant melanoma metastasis. Redox Biol. 2023;60:102619. [ DOI ] [ PMC free article ] [ PubMed ] [ Google Scholar ] 15. Bufka J, Vaňková L, Sýkora J, Křížková V. Exploring carotenoids: metabolism, antioxidants, and impacts on human health. J Funct Foods. 2024;118:106284. [ Google Scholar ] 16. Salles C. Acids in foods and perception of sourness. In: Handbook of Molecular Gastronomy. CRC Press; 2021. p. 7–12. 17. Xu Y, Li P, Yang Z, Xu C. Genetic mapping of quantitative trait loci in crops. Crop J. 2017;5:175–84. [ Google Scholar ] 18. Mauricio R. Mapping quantitative trait loci in plants: uses and caveats for evolutionary biology. Nat Rev Genet. 2001;2:370–81. [ DOI ] [ PubMed ] [ Google Scholar ] 19. Li H, Ribaut JM, Li Z, Wang J. Inclusive composite interval mapping (ICIM) for digenic epistasis of quantitative traits in biparental populations. Theor Appl Genet. 2008;116:243–60. [ DOI ] [ PubMed ] [ Google Scholar ] 20. Li H, Ye G, Wang J. A modified algorithm for the improvement of composite interval mapping. Genetics. 2007;175:361–74. [ DOI ] [ PMC free article ] [ PubMed ] [ Google Scholar ] 21. Li H, Hearne S, Bänziger M, Li Z, Wang J. Statistical properties of QTL linkage mapping in biparental genetic populations. Heredity. 2010;105:257–67. [ DOI ] [ PubMed ] [ Google Scholar ] 22. Lee J. Development and evolution of molecular markers and genetic maps in Capsicum species. In: The Capsicum genome. Cham: Springer International Publishing; 2019. p. 85–103. 23. Chaim A, Borovsky Y, De Jong W, Paran I. Linkage of the A locus for the presence of anthocyanin and fs10. 1, a major fruit-shape QTL in pepper. Theor Appl Genet. 2003;106(5):889–94. [ DOI ] [ PubMed ] [ Google Scholar ] 24. Zygier S, Chaim AB, Efrati A, Kaluzky G, Borovsky Y, Paran I. QTLs mapping for fruit size and shape in chromosomes 2 and 4 in pepper and a comparison of the pepper QTL map with that of tomato. Theor Appl Genet. 2005;111:437–45. [ DOI ] [ PubMed ] [ Google Scholar ] 25. Barchi L, Lefebvre V, Sage-Palloix AM, Lanteri S, Palloix A. QTL analysis of plant development and fruit traits in pepper and performance of selective phenotyping. Theor Appl Genet. 2009;118:1157–71. [ DOI ] [ PubMed ] [ Google Scholar ] 26. Borovsky Y, Paran I. Characterization of fs10. 1, a major QTL controlling fruit elongation in Capsicum . Theor Appl Genet. 2011;123:657–65. [ DOI ] [ PubMed ] [ Google Scholar ] 27. Chunthawodtiporn J, Hill T, Stoffel K, Van Deynze A. Quantitative trait loci controlling fruit size and other horticultural traits in bell pepper ( Capsicum annuum ). Plant Genome. 2018;11:160125. 10.3835/plantgenome2016.12.0125. [ DOI ] [ PMC free article ] [ PubMed ] [ Google Scholar ] 28. Han K, Jeong HJ, Yang HB, Kang SM, Kwon JK, Kim S, et al. An ultra-high-density bin map facilitates high-throughput QTL mapping of horticultural traits in pepper ( Capsicum annuum ). DNA Res. 2016. 10.1093/dnares/dsv038. [ DOI ] [ PMC free article ] [ PubMed ] [ Google Scholar ] 29. Hill T, Ashrafi H, Chin-Wo SR, Stoffel K, Truco MJ, Kozik A, et al. Ultra-high density, transcript-based genetic maps of pepper define recombination in the genome and synteny among related species. G3 (Bethesda Md). 2015;5:2341–55. [ DOI ] [ PMC free article ] [ PubMed ] [ Google Scholar ] 30. Lee HY, Ro NY, Patil A, Lee JH, Kwon JK, Kang BC. Uncovering candidate genes controlling major fruit-related traits in pepper via genotype-by-sequencing based QTL mapping and genome-wide association study. Front Plant Sci. 2020;11:1100. [ DOI ] [ PMC free article ] [ PubMed ] [ Google Scholar ] 31. von Steimker J, Tripodi P, Wendenburg R, Tringovska I, Nankar AN, Stoeva V, et al. The genetic architecture of the pepper metabolome and the biosynthesis of its signature capsianoside metabolites. Curr Biol. 2024;34:4209–23. [ DOI ] [ PubMed ] [ Google Scholar ] 32. von Steimker J, Wendenburg R, Klemmer A, Rosaria M, Fernie AR, Alseekh S, et al. Genome-wide association analysis and linkage mapping decipher the genetic control of primary metabolites and quality traits in Capsicum . Plant J. 2025;122:e70300. [ DOI ] [ PMC free article ] [ PubMed ] [ Google Scholar ] 33. Peterson BK, Weber JN, Kay EH, Fisher HS, Hoekstra HE. Double digest RADseq: an inexpensive method for de novo SNP discovery and genotyping in model and non-model species. PLoS One. 2012. 10.1371/journal.pone.0037135. [ DOI ] [ PMC free article ] [ PubMed ] [ Google Scholar ] 34. Yol E, Basak M, Kızıl S, Lucas SJ, Uzun B. A high-density SNP genetic map construction using ddRAD-seq and mapping of capsule shattering trait in sesame. Front Plant Sci. 2021;12:679659. [ DOI ] [ PMC free article ] [ PubMed ] [ Google Scholar ] 35. Mohan V, Paran I. Molecular Mapping and Identification of QTLs and Genes for Economically Important Traits in the Capsicum Genome. In: Ramchiary N, Kole C, editors. The Capsicum Genome. Compendium of Plant Genomes. Cham: Springer; 2019. 10.1007/978-3-319-97217-6_6. 36. Gonzalo MJ, Brewer MT, Anderson C, Sullivan D, Gray S, van der Knaap E. Tomato fruit shape analysis using morphometric and morphology attributes implemented in Tomato Analyzer software program. J Am Soc Hortic Sci. 2009;134:77–87. [ Google Scholar ] 37. Rosa-Martínez E, Garcìa-Martínez MD, Adalid-Martínez AM, Pereira-Dias L, Casanova C, Soler E, et al. Fruit composition profile of pepper, tomato and eggplant varieties grown under uniform conditions. Food Res Int. 2021;147:110531. [ DOI ] [ PubMed ] [ Google Scholar ] 38. Liu F, Zhao J, Sun H, Xiong C, Sun X, Wang X, et al. Genomes of cultivated and wild Capsicum species provide insights into pepper domestication and population differentiation. Nat Comm. 2023;14:5487. [ DOI ] [ PMC free article ] [ PubMed ] [ Google Scholar ] 39. Li H, Durbin R. Fast and accurate long-read alignment with Burrows–Wheeler transform. Bioinformatics. 2010;26:589–95. [ DOI ] [ PMC free article ] [ PubMed ] [ Google Scholar ] 40. Catchen J, Hohenlohe PA, Bassham S, Amores A, Cresko WA. Stacks: an analysis tool set for population genomics. Mol Ecol. 2013;11:3124–40. [ DOI ] [ PMC free article ] [ PubMed ] [ Google Scholar ] 41. Van Ooijen J. Software for the calculation of genetic linkage maps in experimental populations Kyazma BV. Wageningen, Netherlands. 2006. 42. Holtz Y, David JL, Ranwez V. The genetic map comparator: a user-friendly application to display and compare genetic maps. Bioinformatics. 2017;33:1387–8. [ DOI ] [ PubMed ] [ Google Scholar ] 43. Wickham H. ggplot2: Elegant graphics for data analysis. Springer-Verlag New York; 2016. [ Google Scholar ] 44. Shannon P, Markiel A, Ozier O, Baliga NS, Wang JT, Ramage D, et al. Cytoscape: a software environment for integrated models of biomolecular interaction networks. Genome Res. 2003;13(11):2498–504. [ DOI ] [ PMC free article ] [ PubMed ] [ Google Scholar ] 45. Zhou Y, Zhou B, Pache L, Chang M, Khodabakhshi AH, Tanaseichuk O, et al. Metascape provides a biologist-oriented resource for the analysis of systems-level datasets. Nat Commun. 2019;10(1):1523. [ DOI ] [ PMC free article ] [ PubMed ] [ Google Scholar ] 46. Meng L, Li H, Zhang L, Wang J. QTL icimapping: integrated software for genetic linkage map construction and quantitative trait locus mapping in biparental populations. Crop J. 2015;3:269–83. [ Google Scholar ] 47. Li S, Wang J, Zhang L. Inclusive composite interval mapping of QTL by environment interactions in biparental populations. PLoS One. 2015;10:e0132414. [ DOI ] [ PMC free article ] [ PubMed ] [ Google Scholar ] 48. Pepper Genomics Database http:// ted.bti.cornell.edu/cgi-bin/pepper/index . Accessed 07 January 2026. 49. Jones P, Binns D, Chang HY, Fraser M, Li W, McAnulla C, McWilliam H, Maslen J, Mitchell A, Nuka G, Pesseat S, Quinn AF, Sangrador-Vegas A, Scheremetjew M, Yong SY, Lopez R, Hunter S. InterProScan 5: genome-scale protein function classification. Bioinformatics. 2014;30:1236–40. [ DOI ] [ PMC free article ] [ PubMed ] [ Google Scholar ] 50. UniProt Consortium. UniProt: a hub for protein information. Nucleic Acids Res. 2015;43:D204–12. [ DOI ] [ PMC free article ] [ PubMed ] [ Google Scholar ] 51. Jamann TM, Balint-Kurti PJ, Holland JB. QTL mapping using high-throughput sequencing. Plant Funct genomics: methods protocols, 2015: 257–85. [ DOI ] [ PubMed ] 52. Tripodi P. Next generation sequencing technologies to explore the diversity of germplasm resources: achievements and trends in tomato. Comput Struct Biotechnol J. 2022;20:6250–8. [ DOI ] [ PMC free article ] [ PubMed ] [ Google Scholar ] 53. Hill TA, Chunthawodtiporn J, Ashrafi H, Stoffel K, Weir A, Van Deynze A. Regions underlying population structure and the genomics of organ size determination in Capsicum annuum . Plant Genome. 2017;10:plantgenome2017-03. [ DOI ] [ PubMed ] [ Google Scholar ] 54. Zhang XF, Sun HH, Xu Y, Chen B, Yu SC, Geng SS, et al. Development of a large number of SSR and InDel markers and construction of a high-density genetic map based on a RIL population of pepper ( Capsicum annuum L.). Mol Breed. 2016;36:92. [ Google Scholar ] 55. Zhang XF, Wang GY, Dong TT, Chen B, Du HS, Li CB, et al. High-density genetic map construction and QTL mapping of first flower node in pepper ( Capsicum annuum L). BMC Plant Biol. 2019;19:167. [ DOI ] [ PMC free article ] [ PubMed ] [ Google Scholar ] 56. Zhu Z, Sun B, Wei J, Cai W, Huang Z, Chen C, et al. Construction of a high-density genetic map of an interspecific cross of Capsicum chinense and Capsicum annuum and QTL analysis of floral traits. Sci Rep. 2019;9:1054. [ DOI ] [ PMC free article ] [ PubMed ] [ Google Scholar ] 57. Paran I, van der Voort JR, Lefebvre V, Jahn M, Landry L, van Schriek M, et al. An integrated genetic linkage map of pepper ( Capsicum spp.). Mol Breed. 2004;13:251–61. [ Google Scholar ] 58. Garavello M, Cuenca J, Dreissig S, Fuchs J, Navarro L, Houben A, et al. Analysis of crossover events and allele segregation distortion in interspecific citrus hybrids by single pollen genotyping. Front Plant Sci. 2020;11:61. [ DOI ] [ PMC free article ] [ PubMed ] [ Google Scholar ] 59. Barchi L, Bonnet J, Boudet C, Signoret P, Nagy I, Lanteri S, et al. A high-resolution, intraspecific linkage map of pepper ( Capsicum annuum L.) and selection of reduced recombinant inbred line subsets for fast mapping. Genome. 2007;50:51–60. [ DOI ] [ PubMed ] [ Google Scholar ] 60. Cheng J, Qin C, Tang X, Zhou H, Hu Y, Zhao Z, et al. Development of a SNP array and its application to genetic mapping and diversity assessment in pepper ( Capsicum spp.). Sci Rep. 2016;6:33293. [ DOI ] [ PMC free article ] [ PubMed ] [ Google Scholar ] 61. Bodénès C, Chancerel E, Ehrenmann F, Kremer A, Plomion C. High-density linkage mapping and distribution of segregation distortion regions in the oak genome. DNA Res. 2016;23:115–24. 10.1093/dnares/dsw001. [ DOI ] [ PMC free article ] [ PubMed ] [ Google Scholar ] 62. Kulmuni J, Seifert B, Pamilo P. Segregation distortion causes large-scale differences between male and female genomes in hybrid ants. Proc Natl Acad Sci U S A. 2010;107:7371–6. [ DOI ] [ PMC free article ] [ PubMed ] [ Google Scholar ] 63. Zuo JF, Niu Y, Cheng P, Feng JY, Han SF, Zhang YH, et al. Effect of marker segregation distortion on high density linkage map construction and QTL mapping in Soybean ( Glycine max L). Heredity. 2019;123:579–92. [ DOI ] [ PMC free article ] [ PubMed ] [ Google Scholar ] 64. Li C, Bai G, Chao S, Wang Z. A high-density SNP and SSR consensus map reveals segregation distortion regions in wheat. Biomed Res Int. 2015. 10.1155/2015/830618. [ DOI ] [ PMC free article ] [ PubMed ] [ Google Scholar ] 65. Schauer N, Semel Y, Roessner U, Gur A, Balbo I, Carrari F, Pleban T, Perez-Melis A, Bruedigam C, Kopka J, et al. Comprehensive metabolic profiling and genetic analysis reveal complex genetic control of seed development in tomato. Plant Physiol. 2006;140:1183–97.16607031 [ Google Scholar ] 66. Evers D, Lefèvre I, Legay S, Lamoureux D, Hausman JF, Rosales RG, Herrera MDR, Bonierbale M. Metabolic profiling of potato tubers to identify mQTLs associated with quality traits. Mol Breed. 2012;30:421–35. [ Google Scholar ] 67. Wen W, Li D, Li X, Gao Y, Li W, Li H, Liu J, Liu H, Chen W, Luo J, et al. Metabolome-based genome-wide association study of maize kernel leads to novel loci for metabolite levels. Nat Commun. 2014;5:5138. [ DOI ] [ PMC free article ] [ PubMed ] [ Google Scholar ] 68. Xie F, Wang S, Beavis WD, Xu S. Estimation of genetic variance contributed by a quantitative trait locus: correcting the bias associated with significance tests. Genetics. 2021;219(3):iyab115. [ DOI ] [ PMC free article ] [ PubMed ] [ Google Scholar ] 69. Do THT, Martinoia E, Lee Y. Functions of ABC transporters in plant growth and development. Curr Opin Plant Biol. 2018;41:32–8. [ DOI ] [ PubMed ] [ Google Scholar ] 70. Ofori PA, Mizuno A, Suzuki M, Martinoia E, Reuscher S, Aoki K, Shibata D, Otagaki S, Matsumoto S, Shiratake K. Genome-wide analysis of ATP binding cassette (ABC) transporters in tomato. PLoS ONE. 2018;13:e0200854. [ DOI ] [ PMC free article ] [ PubMed ] [ Google Scholar ] 71. Zhang M, Zhao Y, Nan T, Jiao H, Yue S, Huang L, et al. Genome-wide analysis of Citrus medica ABC transporters reveals the regulation of fruit development by CmABCB19 and CmABCC10 . Plant Physiol Biochem. 2024;215:109027. [ DOI ] [ PubMed ] [ Google Scholar ] 72. Orsi CH, Tanksley SD. Natural variation in an ABC transporter gene associated with seed size evolution in tomato species. PLoS Genet. 2009;5:e1000347. [ DOI ] [ PMC free article ] [ PubMed ] [ Google Scholar ] 73. Lopez-Ortiz C, Dutta SK, Natarajan P, Pena-Garcia Y, Abburi V, Saminathan T, et al. Genome-wide identification and gene expression pattern of ABC transporter gene family in Capsicum spp. PloS one. 2019;14:e0215901. [ DOI ] [ PMC free article ] [ PubMed ] [ Google Scholar ] 74. Colonna V, D’Agostino N, Garrison E, Albrechtsen A, Meisner J, Facchiano A, et al. Genomic diversity and novel genome-wide association with fruit morphology in Capsicum , from 746k polymorphic sites. Sci Rep. 2019;9:10067. [ DOI ] [ PMC free article ] [ PubMed ] [ Google Scholar ] 75. Wang Y, Ma S, Cao X, Li Z, Pan B, Song Y, et al. Morphological, histological and transcriptomic mechanisms underlying different fruit shapes in Capsicum spp. PeerJ. 2024;12:e17909. [ DOI ] [ PMC free article ] [ PubMed ] [ Google Scholar ] 76. An X, Zhao S, Luo X, Chen C, Liu T, Li W, et al. Genome-wide identification and expression analysis of the regulator of chromosome condensation 1 gene family in wheat ( Triticum aestivum L.). Front Plant Sci. 2023;14:1124905. [ DOI ] [ PMC free article ] [ PubMed ] [ Google Scholar ] 77. Qin A, Zhou Y, Ge X, Yu X, Zhao Q, Li C, et al. Overexpression of cotton genes GhHAT5 and GhCRK29 promotes embryonic development in Arabidopsis thaliana. J Cotton Res. 2025;8:33. [ Google Scholar ] 78. Li J, Guo Y, Yang Y. The molecular mechanism of plasma membrane H+-ATPases in plant responses to abiotic stress. J Genet Genom. 2022;49:715–25. [ DOI ] [ PubMed ] [ Google Scholar ] 79. Ben Chaim A, Borovsky E, Rao GU, Tanyolac B, Paran I. fs3.1 : a major fruit shape QTL conserved in Capsicum . Genome. 2003;46:1–9. 10.1139/g02-096. [ DOI ] [ PubMed ] [ Google Scholar ] 80. Yarnes SC, Ashrafi A, Reyes-Chin-Wo S, Hill TA, Stoffel KM, Van Deynze A. Identification of QTLs for capsaicinoids, fruit quality, and plant architecture-related traits in an interspecific Capsicum RIL population. Genome. 2013;56:61–74. 10.1139/gen-2012-0083. [ DOI ] [ PubMed ] [ Google Scholar ] 81. Xu C, Liberatore KL, MacAlister CA, Huang Z, Chu YH, Jiang K, et al. A cascade of arabinosyltransferases controls shoot meristem size in tomato. Nat Genet. 2015;47:784–92. 10.1038/ng.3309. [ DOI ] [ PubMed ] [ Google Scholar ] 82. Borovsky Y, Raz A, Doron-Faigenboim A, Zemach H, Karavani E, Paran I. Pepper fruit elongation is controlled by Capsicum annuum ovate family protein 20. Front Plant Sci. 2022;12:815589. [ DOI ] [ PMC free article ] [ PubMed ] [ Google Scholar ] 83. Cao Y, Zhang K, Yu H, Chen S, Xu D, Zhao H, Cheng F. Pepper variome reveals the history and key loci associated with fruit domestication and diversification. Mol Plant. 2022;11:1744–58. 10.1016/j.molp.2022.09.021. [ DOI ] [ PubMed ] [ Google Scholar ] 84. Shockey J, Browse J. Genome-level and biochemical diversity of the acyl‐activating enzyme superfamily in plants. Plant J. 2011;66:143–60. [ DOI ] [ PubMed ] [ Google Scholar ] 85. Ma X, Qiao YM, Li Y, Yu YN, Gong ZH. Identification of fruit traits related QTLs and a candidate gene, CaBRX , controlling locule number in pepper ( Capsicum annuum L.). Horticulturae. 2022;8:146. [ Google Scholar ] 86. Zhu Y, Zhang S, Yu J. ZmAdSS1 encodes adenylosuccinate synthetase and plays a critical role in maize seed development and the accumulation of nutrients. Plant Sci. 2023;330:111644. [ DOI ] [ PubMed ] [ Google Scholar ] 87. Liu X, Li A, Wang S, Lan C, Wang Y, Li J, et al. Overexpression of Pyrus sinkiangensis HAT5 enhances drought and salt tolerance, and low-temperature sensitivity in transgenic tomato. Front Plant Sci. 2022;13:1036254. [ DOI ] [ PMC free article ] [ PubMed ] [ Google Scholar ] 88. Shi CY, Hussain SB, Guo LX, Yang H, Ning DY, Liu YZ. Genome-wide identification and transcript analysis of vacuolar-ATPase genes in citrus reveal their possible involvement in citrate accumulation. Phytochemistry. 2018;155:147–54. [ DOI ] [ PubMed ] [ Google Scholar ] 89. Yang S, Li K, Peng M, Wang H, Lu J, Cai G, et al. Glutathione metabolism contributes to citric acid tolerance and antioxidant capacity in Acetobacter tropicalis . Food Microbiol. 2025;125:104657. [ DOI ] [ PubMed ] [ Google Scholar ] 90. Reinert S. Quantitative genetics of pleiotropy and its potential for plant sciences. J Plant Physiol. 2022;276:153784. [ DOI ] [ PubMed ] [ Google Scholar ] Associated Data This section collects any data citations, data availability statements, or supplementary materials included in this article. Supplementary Materials Supplementary Material 1. (614.3KB, pdf) Supplementary Material 2. (139.1KB, xlsx) Data Availability Statement The data have been uploaded to the National Center for Biotechnology Information (NCBI) with the BioProject accession PRJNA1403783 ([https://www.ncbi.nlm.nih.gov/bioproject/PRJNA1403783](https:/www.ncbi.nlm.nih.gov/bioproject/PRJNA1403783)) with individual accessions listed in Supplementary Table 3. The VCF file containing raw SNP data of the two mapping populations is available in Zenodo [https://zenodo.org/records/18133695](https:/zenodo.org/records/18133695) (doi: [10.5281/zenodo.18133695](https:/doi.org/10.5281/zenodo.18133695)). All the other data supporting our findings are available within the manuscript and the supplementary data. Articles from BMC Plant Biology are provided here courtesy of BMC ACTIONS View on publisher site PDF (3.1 MB) Cite Collections Permalink PERMALINK Copy RESOURCES Similar articles Cited by other articles Links to NCBI Databases Cite Copy Download .nbib .nbib Format: AMA APA MLA NLM Add to Collections Create a new collection Add to an existing collection Name your collection * Choose a collection Unable to load your collection due to an error Please try again Add Cancel Follow NCBI NCBI on X (formerly known as Twitter) NCBI on Facebook NCBI on LinkedIn NCBI on GitHub NCBI RSS feed Connect with NLM NLM on X (formerly known as Twitter) NLM on Facebook NLM on YouTube National Library of Medicine 8600 Rockville Pike Bethesda, MD 20894 Web Policies FOIA HHS Vulnerability Disclosure Help Accessibility Careers NLM NIH HHS USA.gov Back to Top