Prioritization of Deleterious Mutations Improves Genomic Prediction and Increases the Rate of Genetic Gain in Common Bean ( Phaseolus vulgaris L. ), a Simulation Study

Prioritization of Deleterious Mutations Improves Genomic Prediction and Increases the Rate of Genetic Gain in Common Bean (Phaseolus vulgaris L.), a Simulation Study | bioRxiv /* */ /* */ <!-- <!-- /*! * yepnope1.5.4 * (c) WTFPL, GPLv2 */ (function(a,b,c){function d(a){return"[object Function]"==o.call(a)}function e(a){return"string"==typeof a}function f(){}function g(a){return!a||"loaded"==a||"complete"==a||"uninitialized"==a}function h(){var a=p.shift();q=1,a?a.t?m(function(){("c"==a.t?B.injectCss:B.injectJs)(a.s,0,a.a,a.x,a.e,1)},0):(a(),h()):q=0}function i(a,c,d,e,f,i,j){function k(b){if(!o&&g(l.readyState)&&(u.r=o=1,!q&&h(),l.onload=l.onreadystatechange=null,b)){"img"!=a&&m(function(){t.removeChild(l)},50);for(var d in y[c])y[c].hasOwnProperty(d)&&y[c][d].onload()}}var j=j||B.errorTimeout,l=b.createElement(a),o=0,r=0,u={t:d,s:c,e:f,a:i,x:j};1===y[c]&&(r=1,y[c]=[]),"object"==a?l.data=c:(l.src=c,l.type=a),l.width=l.height="0",l.onerror=l.onload=l.onreadystatechange=function(){k.call(this,r)},p.splice(e,0,u),"img"!=a&&(r||2===y[c]?(t.insertBefore(l,s?null:n),m(k,j)):y[c].push(l))}function j(a,b,c,d,f){return q=0,b=b||"j",e(a)?i("c"==b?v:u,a,b,this.i++,c,d,f):(p.splice(this.i++,0,a),1==p.length&&h()),this}function k(){var a=B;return a.loader={load:j,i:0},a}var l=b.documentElement,m=a.setTimeout,n=b.getElementsByTagName("script")[0],o={}.toString,p=[],q=0,r="MozAppearance"in l.style,s=r&&!!b.createRange().compareNode,t=s?l:n.parentNode,l=a.opera&&"[object Opera]"==o.call(a.opera),l=!!b.attachEvent&&!l,u=r?"object":l?"script":"img",v=l?"script":u,w=Array.isArray||function(a){return"[object Array]"==o.call(a)},x=[],y={},z={timeout:function(a,b){return b.length&&(a.timeout=b[0]),a}},A,B;B=function(a){function b(a){var a=a.split("!"),b=x.length,c=a.pop(),d=a.length,c={url:c,origUrl:c,prefixes:a},e,f,g;for(f=0;f<d;f++)g=a[f].split("="),(e=z[g.shift()])&&(c=e(c,g));for(f=0;f<b;f++)c=x[f](c);return c}function g(a,e,f,g,h){var i=b(a),j=i.autoCallback;i.url.split(".").pop().split("?").shift(),i.bypass||(e&&(e=d(e)?e:e[a]||e[g]||e[a.split("/").pop().split("?")[0]]),i.instead?i.instead(a,e,f,g,h):(y[i.url]?i.noexec=!0:y[i.url]=1,f.load(i.url,i.forceCSS||!i.forceJS&&"css"==i.url.split(".").pop().split("?").shift()?"c":c,i.noexec,i.attrs,i.timeout),(d(e)||d(j))&&f.load(function(){k(),e&&e(i.origUrl,h,g),j&&j(i.origUrl,h,g),y[i.url]=2})))}function h(a,b){function c(a,c){if(a){if(e(a))c||(j=function(){var a=[].slice.call(arguments);k.apply(this,a),l()}),g(a,j,b,0,h);else if(Object(a)===a)for(n in m=function(){var b=0,c;for(c in a)a.hasOwnProperty(c)&&b++;return b}(),a)a.hasOwnProperty(n)&&(!c&&!--m&&(d(j)?j=function(){var a=[].slice.call(arguments);k.apply(this,a),l()}:j[n]=function(a){return function(){var b=[].slice.call(arguments);a&&a.apply(this,b),l()}}(k[n])),g(a[n],j,b,n,h))}else!c&&l()}var h=!!a.test,i=a.load||a.both,j=a.callback||f,k=j,l=a.complete||f,m,n;c(h?a.yep:a.nope,!!i),i&&c(i)}var i,j,l=this.yepnope.loader;if(e(a))g(a,0,l,0);else if(w(a))for(i=0;i (function(w,d,s,l,i){w[l]=w[l]||[];w[l].push({'gtm.start':new Date().getTime(),event:'gtm.js'});var f=d.getElementsByTagName(s)[0];var j=d.createElement(s);var dl=l!='dataLayer'?'&l='+l:'';j.src='//www.googletagmanager.com/gtm.js?id='+i+dl;j.type='text/javascript';j.async=true;f.parentNode.insertBefore(j,f);})(window,document,'script','dataLayer','GTM-M677548'); Skip to main content Home About Submit ALERTS / RSS Search for this keyword Advanced Search New Results Prioritization of Deleterious Mutations Improves Genomic Prediction and Increases the Rate of Genetic Gain in Common Bean ( Phaseolus vulgaris L. ), a Simulation Study View ORCID Profile H. Cordoba-Novoa , View ORCID Profile V. Hoyos-Villegas doi: https://doi.org/10.1101/2025.05.05.652208 H. Cordoba-Novoa 1 McGill University, Department of Plant Science , Montreal, Canada Find this author on Google Scholar Find this author on PubMed Search for this author on this site ORCID record for H. Cordoba-Novoa V. Hoyos-Villegas 1 McGill University, Department of Plant Science , Montreal, Canada 2 Michigan State University, Department of Plant, Soil and Microbial Sciences , 1066 Bogue St, East Lansing, MI, USA Find this author on Google Scholar Find this author on PubMed Search for this author on this site ORCID record for V. Hoyos-Villegas For correspondence: hoyosval{at}msu.edu Abstract Full Text Info/History Metrics Preview PDF Abstract The study of mutations is fundamental to understanding evolution, domestication, and genetics. Characterizing mutations has the potential to accelerate breeding programs through selection and purging of deleterious mutations (DelMut). Here, we investigated how predicting DelMut in breeding populations can improve genomic prediction (GP) and inform strategies to increase the rate of genetic gain. DelMut were annotated in three independent common bean populations using a previously developed random forest (RF) model incorporating phylogenetic and protein information. Deleterious scores from the RF model were mostly around 0.25, with the top 1% ( highly DelMut) of variants scoring between 0.78 – 0.82 among populations. All populations showed variation in the number of highly DelMut per line (max. 13 – 197) and in genetic load. We assessed the impact of incorporating a priori information for variant prioritization and weighting based on predicted deleteriousness in GP models for yield and flowering time. Stochastic simulations were conducted to evaluate how different mating schemes based variable numbers of DelMut per parent affect genetic gain. Variants with higher predicted scores had significantly different effect distributions compared to random or lower-scored markers. Yield predictions were 4.47–12.3% more accurate when markers were weighted by effect and deleterious score; no consistent improvement was observed for flowering time. Simulated breeding cycles showed that selecting parents with fewer highly DelMut consistently increases the rate of genetic gain. These results highlight the potential of DelMut information for variant prioritization and the optimization of common bean breeding programs. The approaches we developed can be assessed in other species to improve the efficacy of crop improvement. Key messages - Predicted deleterious mutations have different distributions of effects based on population composition. - Variant prioritization and differential weighing of markers based on effects and deleterious scores can improve the prediction of yield. - Favoring mating schemes between parents with fewer highly deleterious mutations can increase the rate of genetic gain. Introduction The study of mutations plays a paramount role in understanding the evolution and adaptation of species, as well as genetic and phenotypic variation ( Charlesworth et al., 1993 ; Gossmann et al., 2012 ; Piganeau and Eyre-Walker, 2003 ). Depending on their effect, mutations can be neutral, advantageous or deleterious. The distribution of the effects of mutations helps explain the proportion of mutations that can be beneficial or detrimental depending on environmental forces ( Eyre-Walker and Keightley, 2007 ; Krasovec et al., 2016 ). Nonsynonymous mutations (both advantageous or deleterious) are under strong selective pressures. Strongly advantageous mutations are usually rare ( Eyre-Walker, 2006 ) and highly deleterious mutations (DelMut) are purged (Crow, 1970). However, mildly recessive DelMut with subtle effects can escape purifying selection and are accumulated, known as the genomic genetic load ( Glémin, 2003 ). Multiple factors affect the accumulation of mildly DelMut. In crop species, domestication and selection bottlenecks contribute to the increase of inbreeding in populations, which in turn reduces the effective population size and effectiveness of purifying selection, leading to the accumulation of weakly DelMut ( Charlesworth and Charlesworth, 1999 ; Kono et al., 2016 ). The accumulation patterns and potential role of DelMut in phenotypic variation are topics of particular interest in crop improvement. Studies on the characterization of DelMut in diverse populations have been conducted in several species with varying accumulation patterns depending on the mating system and breeding methods ( Dwivedi et al., 2023 ). In self-pollinated species such as common bean and soybean, breeding has reduced the number of DelMut, but mildly DelMut are fixed and accumulated ( Cordoba-Novoa et al., 2025 ; Kim et al., 2021 ). The accumulation of DelMut can limit selection and hinder the genetic gains in breeding programs ( Moyers et al., 2018 ; Zhu et al., 2022 ). The identification and characterization of DelMut in crop species raises the question of how genetic gain in breeding programs can be further accelerated ( Johnsson et al., 2019 ; Wallace et al., 2018 ). The targeted removal of DelMut using new genomic techniques (NGT) such as genome editing is one of the avenues for crop improvement ( Gao, 2021 ; Glaus et al., 2025 ; Johnsson et al., 2019 ). Other approaches aim to enhance the prediction ability (PA) of genomic selection (GS) models. Some studies have considered the use of predicted DelMut to subset markers for GS or their inclusion as fixed effects in the model ( Valluru et al., 2019 ; Wu et al., 2023 ). Methods that modify the relative importance of markers in the model through the inclusion of posterior variances or the modification of genomic relationship matrices (GRM) have also been considered ( Edwards et al., 2016 ; Long et al., 2023 ; Yang et al., 2017 ). The different approaches have shown varying levels of efficacy depending on the trait, the populations, and the environments. In breeding programs, decisions and final results are simultaneously influenced by various parameters. Stochastic simulations have been adopted as tools to simulate breeding scenarios and probable outcomes for optimizing breeding programs ( Covarrubias-Pazaran et al., 2022 ; Hassanpour et al., 2023 ). Such simulations have the advantages of recreating entire populations with genotypic and phenotypic data at the individual level, which provides precise predictions of the consequences of proposed changes at different stages, such as crossing, evaluation, and selection ( Liu et al., 2019 ; Vieira et al., 2025 ). Stochastic simulations have been employed for the simulation and optimization of animal ( Gorjanc et al., 2018 ; Johnsson et al., 2019 ), and plant breeding programs, including rice ( Fritsche-Neto et al., 2024 ), sweet corn ( Peixoto et al., 2024 ), soybean ( Silva et al., 2021 ), and common bean ( Chiaravallotti et al., 2024 ; Lin et al., 2023 ) considering the adoption of GP frameworks. As a predominantly self-pollinating species, common bean may accumulate DelMut with varying effects. As mildly DelMut can accumulate in different genomic regions, the purging of DelMut from breeding populations and elite material has the potential to increase the effectiveness of breeding programs. However, no studies on the potential practical applications of the inclusion of DelMut in plant breeding pipelines have been conducted. Thus, we investigated the potential impact of using information derived from the prediction of DelMut for optimizing breeding programs. Our objective was to predict DelMut in different common bean breeding populations and evaluate how genomic prediction (GP) could benefit from variant prioritization in the model. Additionally, we hypothesize that selecting parents and designing mating schemes based on a priori information about the presence of DelMut could accelerate the rate of genetic gain over time in breeding programs. Materials and Methods Phenotypic and genotypic data Three independent common bean breeding populations with publicly available phenotypic and genotypic data were used in this study. The populations have different genetic backgrounds, breeding histories, and genotypic datasets, and have been evaluated in different environments. The black bean MAGIC population was developed at McGill University, Canada, with eight parents from the Middle American Diversity (MDP) panel ( Moghaddam et al., 2016 ). Founders were selected to maximize allelic diversity and crossed in a multi-funnel scheme (half-diallel). A total of 18 families and 532 recombinant inbred lines (RIL) were advanced. Field experiments were conducted in two locations in Canada in 2024 ( Cordoba-Novoa et al., 2025 ). For the present study, the evaluations from the Sainte Anne de Bellevue, QC (SAB) location were used based on data quality and the experimental design. Best Linear Unbiased Estimators (BLUE) were calculated fitting a mixed-effect model in the lme4 R package ( Bates et al., 2015 ), and a Complete randomized block design (CRBD)Click or tap here to enter text., and means were corrected for further analysis. Flowering was evaluated as the number of days from planting until 50% of the plot plants had at least 50% of their flowers open (DTF). Yield was recorded as Kg/Ha based on the plot weight. The MAGIC RILs were skim-sequenced using 150 paired-end PCR-free libraries in the Illumina platform and imputed using a Practical Haplotype Graph (PHG) built from Nanopore and Illumina Reads from the eight founders ( Bradbury et al., 2022 ; Cordoba-Novoa et al., 2025 ). A total of 1.5 million SNPs were available for the present study. The second population was the Vivero Equipo Frijol (VEF). VEF is an elite Andean population that is part of the bean breeding program of the International Center of Tropical Agriculture (CIAT). Details on the population and phenotypic evaluation can be found in Keller et al. (2020) . For the present study, we used the VEF phenotypic data for 346 genotypes evaluated in Darien (Valle del Cauca, Colombia) under the middle phosphorous level (DAR16C_mdP location). The location was selected among the locations with the greatest number of evaluated genotypes and with similar prediction abilities in the environment (DAR16C) where VEF was evaluated. BLUE-corrected means for DTF and yield (Kg/Ha) were available and directly used for our analysis. As previously reported, the VEF population was genotyped using Genotype-by-sequencing (GBS) with the ApeKI restriction enzyme and Illumina sequencing. After quality control, 5,820 SNP markers were available. The third population was the Cooperative Dry Bean Nursery (CDBN), a multi-environment trial (MET) grown for over 70 years in the US and Canada. MacQueen et al. (2020) reported the data analysis strategies for four decades of the CDBN evaluation. For yield, we used phenotypic data from 318 genotypes evaluated at the MSTI (Sidney, Montana, US) location, and for DTF, we used data from 226 genotypes evaluated at the WYPO (Powell, Wyoming, US) location. These locations had the highest number of observations (datapoints) for each trait. Due to the unbalanced nature of the data, Best Linear Unbiased Predictors (BLUP) were calculated across years. MacQueen et al. (2020) genotyped the CDBN accessions by re-analyzing raw sequencing from previous reports and sequencing some genotypes with a dual-enzyme ( MseI and TaqI ) GBS approach. After variant calling and QC, a total of 1.2 million SNPs were available. Population structure and LD analyses As mentioned above, the populations used were previously reported and characterized to a certain extent. To facilitate the analysis of the results, we evaluated population structure using principal component analysis (PCA) implemented in Tassel v5 ( Bradbury et al., 2007 ) and pairwise linkage disequilibrium (LD) in GAPIT v3 ( Wang and Zhang, 2021 ). LD was plotted using the R package LDheatmap ( Shin et al., 2006 ). For population structure and LD evaluation, the genotypic data sets of the MAGIC population and CDBN were reduced to a similar number of markers by setting a minimum distance of 10 Kbp between SNPs while maintaining the LD patterns. Prediction of putatively deleterious mutations To identify putatively DelMut, two Random Forest (RF) models for common bean were reported by Cordoba-Novoa et al. (2025) , one trained based on the Middle American and another on the Andean Common bean genome reference. The Middle American RF model was used for the MAGIC population, whereas the Andean RF model was used for the VEF and the CDBN. The model choice depends on the genetic background and the reference genome (UI111 – Middle American or G19833 – Andean) used for variant calling in each population. Briefly, for the model implementation, Sorting Intolerant From Tolerant For Genomes (SIFT4G; Ng and Henikoff, 2003 ) scores and unified representation (UniRep) embeddings ( Alley et al., 2019 ) are obtained for each genotypic dataset and later loaded in the RF models to calculate a deleterious score. The higher the value from zero to one, the more deleterious an allelic change is predicted to be. The SNPs in the top 1% of the deleterious scores within each population were classified as highly deleterious ( highly DelMut). To obtain a genome-wide estimation of the deleterious burden of each genotype in each population, the homozygous (Hom), heterozygous (Het), and total genetic load per genotype was estimated as the sum of all the predicted scores for a line according to the allelic state (i.e. if the derived allele is not present (or is the reference, the score is not added) following a similar approach as Wu et al. (2023) . Estimation of effect sizes Following the approach proposed by Valluru et al. (2019) the effect size of the SNPs was estimated using a RR-BLUP model in the R-package rrBLUP v4.6 ( Endelman, 2011 ). Where y is the vector of BLUEs or BLUPs for yield or DTF, W is the matrix relating individuals to observations, G is the genotype matrix coded as {-1,0,1} under an additive model, and u ∼ N (0, I ) is the vector of marker effects. The marker effects can be written as û = Z′ ( ZZ′ + λI ) −1 y , where Z=WG and λ is the ridge regression parameter defined as the ratio between residual and marker variances ( / ). For each population, the distributions of the effects of SNPs with and without score were compared. Since the number of SNPs with scores is lower, random subsets of equal size from SNPs with no scores were defined. The minor allele frequency (MAF) was verified among subsets to avoid bias. SNPs Scored for DelMuts were further subset, and effects re-calculated to compare the effect distributions between the SNPs with the top 30% DelMut score and the remaining 70% based on weight distribution and percentiles. Differences between distributions were evaluated using the Kolmogorov–Smirnov (Ks) test in R. Genomic prediction To evaluate the potential of incorporating information from DelMut into genomic prediction (GP), a Genomic Best Linear Unbiased Prediction (GBLUP) and a Bayesian Ridge Regression (BRR, also known as Gaussian prior) model were used. The following GBLUP model was implemented in ASReml (Butler et al., 2023). Where y is the vector of phenotypic BLUEs or BLUPs, X is a design matrix relating the fixed effects to each genotype, b is the vector of fixed effects, Z is a design allocating the records of genetic values, u is the vector of additive genetic effects for a genotype, and e is the vector of random normal errors. In the model, var ( u ) = where G is the genomic relationship matrix (GRM) and is the genetic variance for the model. The GRM was constructed using the additive method from VanRaden (2008) in the R package AGHmatrix ( Amadeu et al., 2023 ) as follows: Where M is the centered genotype matrix coded as 0, 1 and 2 for the homozygous reference allele, heterozygous, and homozygous alternative allele, respectively. D is the identity matrix, and p and q are the allele frequencies. The following BRR model was implemented in the BGLR package ( Pérez and De Los Campos, 2014 ). Where X is the design matrix, and e is the independent and identical distributed Gaussian error. w is the vector of marker effects (model parameters) where w ∼ N (0, λ −1 I). In the model, y is the vector of phenotypic values with . In the BRR model, the number of Markov Chain Monte–Carlo iterations per model was 12,000 with the first 2,000 as burn-in without partition of training and testing datasets. To evaluate model performance, each population was randomly split into 70% and 30% for the training and testing datasets. The predictions were compared based on the results from a 10-fold cross-validation with 50 iterations for GBLUP and 10 for BRR. The average prediction ability (PA) was determined by Pearson’s correlation between the predicted and the observed phenotypic values. Mean PAs were compared using Student’s t-test. Inclusion of deleterious mutations information in Genomic Prediction For the inclusion of DelMut information in the GBLUP model, new GRMs were constructed weighing the SNPs based on their effect and predicted DelMut scores. The SNP weights are the absolute value of the estimated effect (from the rrBLUP model) times the predicted score: SNPs with the same or similar estimated effect will be considered differently in the model depending on how deleterious (expressed by the score) the allelic change is predicted to be. SNPs with no DelMut scores were kept unweighted in the model. The GRMs were weighed according to Liu et al. (2020) for the VanRaden method used. Unlike GBLUP, the BRR model starts from the genotype matrix rather than a GRM. To inform the model about DelMut, the marker dataset was split into scored and unscored SNPs and defined as two different random effects in a two-level list in the model, following a multi-kernel approach ( Pérez and De Los Campos, 2014 ). To avoid potential size bias, in each iteration the SNP subsets were randomly sampled and kept to the same number of markers. Genomic relationship matrix dimension reduction For the VEF population, no further filtering or reduction was applied to the SNP dataset. Based on the original GP study for VEF that showed that the marker number (higher than 1K markers) did not affect the model PA, and to give enough room to accommodate SNPs with deleterious scores, subsets of 3K SNPs were used for the iterations ( Keller et al., 2020 ). Due to the high number of SNPs and to alleviate computational burden, the MAGIC and CDBN genotypic datasets were reduced using the --indep-pairwise tool in Plink v1.9 ( Chang et al., 2015 ) with a window size of 50Kb, step of 1 and a r 2 threshold of 0.3, similar to previous reports for common bean ( Keller et al., 2020 ). After reducing the MAGIC and CDBN marker datasets, SNPs with scores were re-incorporated into the baseline dataset if they had been deleted during the pruning step. In each iteration of these two populations, we used subsets of 7K randomly selected markers, maintaining an equal number of SNPs with and without deleterious scores (weighted and unweighted, respectively). Simulation of the genetic gain We evaluated how incorporating information on DelMut could improve selection and increase the rate of genetic gain over time. As proof of concept, we simulated a hypothetical basic breeding scenario using a pedigree breeding strategy with artificial selection for a quantitative trait in each generation (represented in Lin et al., 2023 ). Here, the initial crosses of the first breeding cycle are carried out between parents solely selected based on the number of highly DelMut. Three independent crossing schemes (closed programs) were compared. First, genotypes with a high number of highly DelMut were crossed between them High × High ( H × H ); second, parents with a low number of highly DelMut were crossed Low × Low ( L × L ); and third, parents with a high number of highly DelMut were crossed with parents with a low number of highly DelMut High × Low ( H × L ). An additional scheme of random crosses was also simulated. Within each proposed scenario, the top (higher number of DelMut) and the bottom (lower number of DelMut) 20 parents were used for the initial random crosses. For subsequent breeding cycles (second onwards), parents are selected from the advanced lines of the previous cycle based on the phenotype (top simulated parents are included in the next cycle). After cycle 0 was complete, no further selection based on DelMut content was made among the set of parents that made up cycle 1 onwards. In the simulations, 10 breeding cycles from initial crosses to advance yield trials (AYT) were simulated with 50 iterations in AlphaSimR ( Chris Gaynor et al., 2021 ). Real (not simulated) genotypic and phenotypic data collected for each population (VEF, MAGIC, CDBN) were used for the simulations. Average genetic gain was calculated per cycle as the difference between the mean genetic value of the varieties and mean genetic value of the parents in each cycle. Cumulative genetic gain was calculated in each cycle by adding the previous cycle’s shift in genotypic value. To keep comparisons stable among populations, broad-sense heritability (H 2 ) in AlphaSimR was set as 0.25 for yield and 0.45 for flowering based on previous reports for common bean ( Delfini et al., 2021 ; Kamfwa et al., 2015 ; Keller et al., 2020 ; Raggi et al., 2019 ). The genetic map reported by Diaz et al. (2020) was used for simulations. Results Population structure and LD The three populations, VEF, MAGIC, and CDBN, have different breeding histories, genetic backgrounds and were genotyped using various methods, which makes direct comparisons challenging. However, different population structures and LD patterns were observed within each population. In the PCA, VEF had a main cluster with a few lines on the far right and explained variation of 22.4% between PC1 and PC2 ( Figure 1A ). The MAGIC population had two main subpopulations, mainly based on the PC1, that explained 9.5% of the variation ( Figure 1B ). The CDBN had three clear subpopulations tightly clustered with PC1 and PC2, explaining up to 52.4% of the variation ( Figure 1C ). Download figure Open in new tab Figure 1. Principal component analysis (PCA) for population structure in Vivero Equipo Frijol (VEF; A), Black bean MAGIC population (B), and Cooperative Dry Bean Nursery (CDBN; C). Linkage disequilibrium (LD) heatmaps for chromosome 01 ( Pv 01) in VEF (D), MAGIC (E), and CDBN (F). The SNP dataset is independent for each population. LD patterns for the other chromosomes in Figure S1-3. Despite differences in marker density, similar chromosome lengths were observed in the LD analysis of each population. Different LD patterns were identified in each population (Figure S1-3). Figure 1 D-F shows the LD heatmap for Pv 01 in each population as an example of the varying patterns observed in each population. For instance, a big cluster of markers in high LD (r 2 > 0.8) was observed in the middle of Pv01 for VEF, while a smaller cluster, still in high LD, was identified at the beginning of the chromosome for MAGIC. In CDBN, no clear and large LD blocks were observed. Phenotypes Descriptions of the phenotypic data can be found in the original reports for each population (Cordoba-Novoa, 2025; Keller et al., 2020 ; MacQueen et al., 2020 ). Briefly, in VEF, yield ranged from 457.5 to 1,815 Kg/Ha with a mean of 1,072.6 Kg/Ha (CV = 22.9%). Flowering ranged between 35.4 to 40.2 days and a mean of 38.2 days (CV = 2.5%). In the MAGIC RILs, yield ranged between 544.7 – 3,177.5 Kg/Ha, with a mean of 1,886.2 Kg/Ha (CV = 20.1%), and DTF from 41.5 to 50.5 days and a mean of 45.7 days (CV = 3.9%). In the CDBN dataset, yield in the MTSI location had higher values compared to VEF and MAGIC, ranging from 2,063.3 to 3,832.8 Kg/Ha and a mean of 3,065 Kg/Ha (CV = 9.9%). In the WYPO location, DTF varied from 54.3 to 62.2 days, with a mean of 58.4 days (CV = 2.3%). Prediction of deleterious mutations in populations The number of annotated SNPs with deleterious scores varied in each population depending on the data availability. For the VEF population, 1,072 SNPs had a predicted score, 4,753 in the MAGIC population, and 14,740 in the CDBN dataset. Scores varied from zero to one, with most of the values around 0.25 in all three populations ( Figure 2A-C ). Due to a higher marker density, in the CDBN there was a wider score distribution. Based on the top 1%, the threshold for highly DelMut was 0.82 for VEF and CDBN, and 0.78 for MAGIC. In the three datasets, the number of highly DelMut was between 1.18 – 1.34% of the SNPs with a predicted score. For the total genetic load, the three populations showed normal distributions with values between 250 – 350 for VEF, 200 – 600 for MAGIC, and 4400 – 5200 for CDBN ( Figure 2D-F ; Table S1-3). Interestingly, in the distribution of the CDBN, only a few lines within the Middle American group (n = 15) had total loads between 4800 and 5000, with two peaks in the distribution ( Figure 2F ; Table S3). The homozygous genetic load calculated only with the highly DelMut showed the same distribution patterns for the three populations (Figure S4). Download figure Open in new tab Figure 2. Prediction of deleterious mutations in three common bean breeding populations, Vivero Equipo Frijol (VEF; A, D), Black bean MAGIC population (B, E), and Cooperative Dry Bean Nursery (CDBN; C, F). Distribution of deleterious scores (A-C) and distribution of total genetic load (D-F). The SNP dataset is independent for each population. Scored SNPs accumulated in 800 genes for the VEF population (Table S4), 7,639 for MAGIC (Table S5), and 4,943 in CDBN (Table S6). Highly DelMut accumulated in 12 genes in VEF, 37 genes in MAGIC, and 143 genes in CDBN with 1-3 mutations per gene (Table S7). Estimation of effect sizes The distributions of the absolute values of SNP effects showed differences between scored and unscored SNPs for yield and DTF. For yield, the effects of unscored SNPs were more evenly distributed compared to the scored SNPs. The effects for scored SNPs were mainly concentrated around zero, with a highly leptokurtic distribution and higher peaks ( Figure 3 ). Significant differences ( p < 0.001) between the distributions were observed in the MAGIC and CBDN populations, with marginal differences in VEF ( p < 0.05). When the scored SNPs were split based on the 30% percentile score threshold (0.39 for VEF, 0.38 for MAGIC, 0.5 for CDBN), a similar pattern was observed (Figure S5). Markers with a higher score (predicted to be more deleterious) had a near zero peak compared to SNPs with lower score values. For DTF, the same behavior for the distribution of scored and unscored SNPs was observed in the three populations ( Figure 4 ). When comparing scored SNPs for the CDBN dataset, opposite to the trend, SNPs with scores 0.5 (Figure S6 E-F). Download figure Open in new tab Figure 3. Distribution of the absolute value of the estimated effect of scored and unscored SNPs for yield in Vivero Equipo Frijol (VEF; A), Black bean MAGIC population (C), and Cooperative Dry Bean Nursery (CDBN; E). Cumulative probability for the distribution of the effects (B-F). D and p-value correspond to the Kolmogorov–Smirnov (Ks) test for distributions. The SNP dataset is independent for each population. Download figure Open in new tab Figure 4. Distribution of the absolute value of the estimated effect of scored and unscored SNPs for days to flowering (DTF) in Vivero Equipo Frijol (VEF; A), Black bean MAGIC population (C), and Cooperative Dry Bean Nursery (CDBN; E). Cumulative probability for the distribution of the effects (B, D, F). D and p-value correspond to the Kolmogorov–Smirnov (Ks) test for distributions. The SNP dataset is independent for each population. Genomic prediction We investigated how incorporating previous information on DelMut could improve the prediction ability of genomic prediction models. For yield, the unweighted GBLUP and simple BRR had similar results with higher prediction abilities in the CDBN dataset (0.50 - 0.51), followed by VEF (0.38 - 0.40), and MAGIC (0.25). When incorporating the information from DelMut (weights based on effects and deleterious score for GBLUP, and multi-kernel for BRR), the prediction ability increased for all three datasets using GBLUP, and for VEF and CDBN using BRR. The weighted GBLUP model resulted in a 12.3% increase in the PA in the VEF population (0.52), 5.45% (0.31) in MAGIC and 4.47% (0.55) in CDBN. The multi-kernel BRR increased the PA by 6.5% in VEF and 3.0% in CDBN ( Figure 5A ). Significant differences were only detected for GBLUP. Download figure Open in new tab Figure 5. Prediction accuracy for yield (A) and days to flowering (B) using weighted and unweighted GBLUP, and simple and multi-kernel BRR models for the Vivero Equipo Frijol (VEF), Black bean MAGIC, and Cooperative Dry Bean Nursery (CDBN) populations. *** indicates significant differences ( p < 0.0001) according to student’s t-test. For flowering (DTF), the unweighted GBLUP had PA of 0.49 for VEF, 0.46 for CDBN, and 0.25 for MAGIC. Similar PAs were obtained for VEF and CDBN using the simple BRR model. However, a higher PA for DTF in MAGIC was obtained using the simple BRR model (0.57). The incorporation of weights in GBLUP marginally improved the PA of the models for DTF in MAGIC (0.88% higher) and CDBN (0.93% higher). The multi-kernel BRR model had similar results to the GBLUP models with a marginal increase of 0.23% in the PA in VEF ( Figure 5B ). Simulation of the genetic gain Genotypes from each population were selected as potential parents to simulate new closed breeding cycles with a pedigree method. Parents were selected solely based on the number of highly DelMut alleles. In the VEF population, parents with a “Low” number had seven to eight mutations and “High” had 12-13 (Table S8). In the MAGIC population, selected parents in the “Low” group had 0-3 mutations, and those in the “High” group had 16-19 (Table S9). From the CDBN population, “Low” DelMut genotypes had 16-19 mutations, and “ High ” genotypes had between 31-33 (Table S10). Random simulated crosses behaved similarly to the H × L scheme, the results were excluded to facilitate visualization and comparison of the defined mating systems based on DelMut. The genetic gain for yield decreased every cycle in each population. In the first breeding cycle of the derived population from VEF, the genetic gain was 1.37 Kg/Ha in the H × H scheme, 2.17 Kg/Ha in the H × L scheme, and 2.22 Kg/Ha in the L × L. In the second cycle, L × L showed a higher genetic gain (1.63 Kg/Ha) compared to the other two strategies (0.39 – 0.67 Kg/Ha). The increase in genetic gain was close to zero after the fifth cycle for H × L and H × H crossing schemes, whereas small increases were observed for the L × L strategy until the seventh cycle ( Figure 6A ). In the cumulative genetic gain, higher increases were observed for the crosses made from parents with a low number of highly DelMut (L × L) with a final gain of 5.85 Kg/Ha (0.54%) followed by the H × L crosses with 4.16 Kg/Ha (0.39%), and finally the H × H crosses with 2.43 Kg/Ha (0.23%). Download figure Open in new tab Figure 6. Simulation of genetic gain per cycle and cumulative genetic gain for yield (A-C) and days to flowering (D-F) for the Vivero Equipo Frijol (VEF; A, D), Black bean MAGIC (B, E), and Cooperative Dry Bean Nursery (CDBN; C, F) populations. High × High, High × Low, and Low × Low indicate the crossing schemes based on the number of highly deleterious mutations in each population. Similar trends in the reduction of genetic gain per cycle and the increase of cumulative genetic gain were observed for all the selection strategies in the MAGIC and CDBN populations for yield. In the MAGIC population, lower genetic gains were predicted ( Figure 6B ). In the first cycle, the yield increase for the L × L scheme was 1.44 Kg/Ha, 1.17 Kg/Ha for H × L, and 1.05 Kg/Ha in H × H. By the 10 th cycle, the cumulative genetic gain was lower in H × H (2.18 Kg/Ha; 0.11%) compared to both L × L (2.49 Kg/Ha; 0.13%) and H × L (2.40 Kg/Ha; 0.13%). In the simulation from the CDBN genotypes, the genetic gain in the first cycle was around 2.06 Kg/Ha for the L × L and the H × H crossing schemes, followed by the H × L with 1.46 Kg/Ha. The final cumulative genetic gain across crossing strategies was lower than VEF (but higher than in MAGIC). As in VEF and MAGIC, the cumulative genetic gain was highest in the L × L scheme, reaching 4.59 Kg/Ha (0.15%). However, the H × H derived crosses had a final genetic gain of 4.09 Kg/Ha (0.13%) followed by the H × L scheme with 3.57 Kh/Ha (0.12%; Figure 6C ). Low genetic gains for flowering were observed in the simulated cycles from the VEF and CDBN populations. In the VEF simulations, H × H and H × L had higher genetic gains in the first cycle (1.72 and 1.5 days, respectively) compared to L × L (1.04 days). By the sixth cycle all schemes no longer showed a genetic gain per cycle ( Figure 6D ). Final cumulative genetic gains were similar for H × H and H × L (3.35 days; 8.5%) and slightly lower for L × L (3.17 days; 8.0%). For the MAGIC and CDBN simulations, clearer differences among crossing schemes were observed. In MAGIC, higher genetic gains for flowering compared to VEF and CDBN were observed. In the first cycle, the H × L crosses gained 2.45 days, H × H 1.53 days, and L × L 1.38 days. This trend between crossing schemes was maintained in the cumulative genetic gain, with a final increase of 5.11 days (10.7%) in H × L, 4.01 days (8.5%) in H × H, and 3.65 days (7.7%) in L × L. For the CDBN simulation, the L × L strategy had lower genetic gains in DTF compared to the other two and reached zero at the fourth cycle. The H × H and H × L schemes registered gains until the sixth cycle ( Figure 6F ). The final cumulative gain in DTF was higher in H × H (3.04 days; 5.1%) and H × L (2.82 days; 4.8%) than in the L × L cycles (1.97 days; 3.3%). Discussion Effect of deleterious mutations We investigated how the prediction of DelMut can potentially inform selection decisions and accelerate breeding programs. We evaluated three publicly available common bean breeding populations with diverse genetic and breeding histories. Methods for the prediction of DelMut based on conservation constraints have been previously developed ( Davydov et al., 2010 ; Ng and Henikoff, 2003 ). The adopted approach has the advantage of combining both phylogenetic information and the likely impact of amino acid substitutions, providing a continuous classification of deleteriousness for variant prioritization ( Cordoba-Novoa et al., 2025 ; Long et al., 2023 ; Ramstein and Buckler, 2022 ). The observed scores had an expected distribution where most of the annotated SNPs have low values around 0.25. Mutations with a low score are likely to be tolerated, while mutations with a high score are predicted to be highly deleterious ( Cordoba-Novoa et al., 2025 ; Long et al., 2023 ). When considered individually, the effects of DelMut may seem inappreciable. It is the aggregated effect of multiple putatively DelMut with subtle individual effects that may have an impact on the overall plant fitness ( Felsenstein, 1974 ; Kono et al., 2019 ). The distribution of the fitness effects is a research topic itself and explains the proportion of mutations that may be deleterious, neutral or advantageous within populations ( Eyre-Walker and Keightley, 2007 ). Mutation accumulation experiments and other studies have shown that the effects of mildly DelMut generally follow a leptokurtic gamma distribution with a shape parameter ( α ) lower than one ( Eyre-Walker et al., 2006 ; Keightley, 1996 ). A highly leptokurtic gamma distribution was observed for the scored SNPs (and above certain threshold) in yield and flowering ( Figure 3 and 4 ). This indicates that most of the mutations are mildly deleterious, and a few of them have large (positive or negative) fitness effects in the long right tail of the distribution ( Keightley, 1996 ). These results are consistent with the distribution of the predicted scores ( Figure 2A-C ) and the observed distributions within scored SNPs (Figure S5-6). Additionally, Ohta and Kimura (1971) suggested that mutations are not equally deleterious but range from near-neutral to slightly deleterious to very deleterious alleles. Leptokurtic distributions with low shape parameters are characterized by the clustering of values at the minimum (Low effect; Böndel et al., 2022 ; Piganeau and Eyre-Walker, 2003 ). In the biological context of DelMut, large-effect mutations that can be lethal or highly detrimental to plant fitness are under strong purifying selection and eventually eliminated from the population. However, small-effect mutations can be tolerated and accumulated at a higher rate. Domestication and breeding reduce effective population sizes, increase inbreeding, and reduce the effectiveness of purifying selection, which leads to the accumulation of tolerated mildly and a few highly DelMut ( Grossen and Ramakrishnan, 2024 ; Kono et al., 2016 ). Mutations with high effects and scores are predicted to have more significant impacts on plant fitness compared to other variants and may be more informative for selection approaches. Genomic prediction The prediction of phenotypes from genotypic data using trained models on related populations has been of extensive interest in plant and animal breeding. Multiple factors affect the PA of models. Strategies such as variant selection and differential marker weighing have been proposed to improve genomic selection efforts. There are multiple factors affecting the prediction accuracy (PA) of GP, such as the training and testing populations, model selection, and variant selection ( Alemu et al., 2024 ; VanRaden et al., 2017 ). GBLUP is a widely used model that has been demonstrated to generally be superior in predicting different traits ( Wang et al., 2018 ). On the other hand, with a Bayesian framework, BRR directly models marker effects, could be more flexible, and has been used for GP in common bean ( de los Campos et al., 2013 ; Keller et al., 2020 ). GBLUP assumes that all markers contribute equally to the total genetic variance, modeling breeding values through a genomic relationship matrix. BRR, in contrast, assumes normally distributed marker effects with equal variance, estimating them directly in a Bayesian framework. Here, we consider the fact that small- and large-effect DelMut can accumulate at different rates in breeding populations and how conservation and predicted deleteriousness can assist variant prioritization to optimize GP. In our results, differentially weighting SNPs based on their effects and predicted deleteriousness scores in GBLUP, as well as partitioning scored and unscored SNPs as separate random effects in the BRR model, improved the prediction accuracy (PA) of GBLUP for yield across all three evaluated populations ( Figure 5A ). VanRaden et al. (2017) and Xavier (2019) highlight the importance of variant selection, pruning, and filtering to enhance genomic prediction. Causal variants and markers closer to them or in strong linkage are expected to explain higher proportions of variance compared to random markers ( Meuwissen et al., 2024 ). Very conservative filtering methods may result in losses in PA, while an excessive number of markers increase computational burdens. Previous reports have shown that weighted GP approaches can improve prediction models in animal and plant species with varying success ( Fang et al., 2017 ; Long et al., 2023 ; Meuwissen et al., 2024 ; Zhang et al., 2024 ). Yield is a highly complex trait controlled by multiple small-effect genes and their interactions. One characteristic of the scored SNPs from the RF model is that they are in coding regions ( Cordoba-Novoa et al., 2025 ; Ramstein and Buckler, 2022 ). Prioritizing putatively DelMut with individual small effects could help the model capture a substantial and biologically relevant cumulative effect ( de los Campos et al., 2013 ; Goddard et al., 2010 ). Another plausible explanation is that prioritized SNPs (scored and weighted) can be in a strong linkage disequilibrium (LD) with unknown causal variants, contributing to a better capture of the genetic variation and prediction ( Vilhjálmsson and Nordborg, 2012 ). Kono et al. (2019) showed in barley that top progeny selected based on an RR-BLUP model had fewer DelMut relative to other lines. This suggests that DelMut affect the observed phenotypes and GP models indirectly account for putatively DelMut. In this regard, in other self-pollinating species such as soybean, elite lines have been shown to have a lower number of DelMut ( Kim et al., 2021 ). Negative correlations between DelMut (and genetic loads) and yield have also been reported for common bean ( Cordoba-Novoa et al., 2025 ), maize ( Yang et al., 2017 ), and potato hybrid populations ( Wu et al., 2023 ). The VEF population was previously evaluated in multiple environments and traits using a genomic prediction framework. The PAs we obtained for yield and flowering with either GBLUP or BRR follow the previous study ( Keller et al., 2020 ). Keller et al. (2020) found that the inclusion of significant QTL signals as fixed effects did not improve the PA of the model. Instead of accounting for fixed effects, our approach directly modifies how SNPs are accounted for in the model based on the effect and/or deleterious score, showing consistent PA improvements in VEF and other populations for yield. Predictions for days to flowering (DTF) did not benefit from incorporating DelMut information (weights or multi-kernel partition) into the model. Flowering time is a less complex trait than yield, with a larger proportion of phenotypic variance attributed to genetic factors (higher broad-sense heritability). As a less complex trait, flowering time is controlled by fewer, bigger-effect genes compared to yield. A higher number of small-effect genes involved in the multiple subprocesses affecting yield also leads to a higher accumulation of DelMut affecting the trait (more genes accumulate more DelMut related to the trait). This may explain why the prediction of yield is more responsive to the inclusion of DelMut information in the GP models. In contrast to yield, significant QTL and genes identified through linkage mapping and GWAS have been reported for DTF ( Ates et al., 2018 ; Nascimento et al., 2018 ; Raggi et al., 2019 ). Given that these QTLs account for a significant percentage of the phenotypic variation, additive genetic effects may already be effectively captured in the model, with no apparent advantages from SNPs with minor effects, such as the scored ones. When examining the effects of marker selection for GP in radiata pine ( Pinus radiata ) and shining gum ( Eucalyptus nitens ), Klápště et al. (2020) also noted that low heritability traits benefit more from model refinement. In simulation studies, Morgante et al. (2018) demonstrated that the prediction of traits governed by a greater number of QTLs (more complex) are more susceptible to parameter tuning compared to traits of lower complexity. Simulation of the genetic gain Stochastic simulations support the optimization of breeding programs under varying parameters. Mating system designs influence the efficiency of breeding cycles and the rate of genetic gain over time. In our simulated closed system, the selection of parents based on the number of highly DelMut influenced the expected genetic gain over time. For yield, higher per-cycle and cumulative genetic gains were consistently observed in all the populations for breeding schemes where the initial parents had a low number of highly DelMut ( L × L ). Since the genetic load of plants does not have evident phenotypic effects and DelMut are not routinely considered in GP schemes, the genome-wide prediction and later consideration of DelMut in breeding materials can inform selection decisions, particularly in the choice of parents. The incorporation of the knowledge about DelMut is of especial importance in animal breeding. For instance, Sonesson et al. (2003) and Raoul et al. (2018) stress the value of selecting against carriers and designing mating systems based on the incorporation or exclusion of these individuals. Carriers are individuals known to carry recessive mutations associated with diseases that could express later in the breeding process. Johnsson et al. (2019) simulated how the selection against carriers and genome editing of DelMut increase animal average fitness. They observed that the most effective strategy depends on the DelMut level of dominance (codominant or recessive). Selection against carries is more effective for recessive mutations, whereas genome editing has the potential for both. Different gene action models have been proposed for DelMut, including additive, incomplete dominance, and codominance ( Robinson et al., 2023 ; Sun et al., 2023 ). In an additive action model, the small effects of the predicted DelMut are independent and add up linearly, which may contribute to the observed improvements in GP and the final simulated genetic gain in each crossing scheme. In other species such as maize, incomplete dominance of DelMut can also contribute to trait variation and heterosis ( Yang et al., 2017 ). Additional research on other gene action effects of DelMut may open new avenues on how effects are modeled and then incorporated in breeding programs. In our cumulative genetic gain of DTF, breeding initiated with parents with a low number of highly DelMut ( L × L ) had a lower gain. Short flowering time has been positively correlated with yield in previous studies ( Kamfwa et al., 2015 ; Moghaddam et al., 2016 ). It is possible that when selecting against parents with a high number of highly DelMut, unknown mutations in genes involved in flowering are indirectly excluded, particularly in the first breeding cycle. For instance, the repair of a domestication DelMut in a floral regulator induced early fruit yield and more compact plants in tomato ( Glaus et al., 2025 ). After the first cycle, parents are selected based on simulated phenotypes (top ones), which, based on the relationship between yield and flowering, can influence a lower flowering time in the cumulative genetic gain. Based on our results, the selection and design of mating (crossing) schemes between parents with a low number of predicted DelMut ( L × L ) has the potential to increase the genetic gain in yield and reduce flowering time. The accumulation of DelMut can limit the rate of genetic gain in breeding programs ( Moyers et al., 2018 ). Domestication and improvement increase inbreeding and reduce the effective population size, which in turns reduces the effectiveness of natural selection and leads to the accumulation of mildly and slightly DelMut ( Dussex et al., 2023 ; Renaut and Rieseberg, 2015 ). Depending on the breeding purpose and the target environment and conditions, the effects of DelMut can be associated with their effect on the overall plant fitness. In a simplified manner, fitness can be defined as the ability of an organism to survive and reproduce in a specific environment ( Orr, 2009 ). The level of fitness of an individual is related to the proportion of the next generation represented by its offspring, and depends on the contextual framework. In a breeding context, for instance, plants able to survive biotic and abiotic stress or yield more in different environments are expected to have a higher fitness. That same ability to produce more can be affected by the genetic load of the individual or the population if analyzed as a whole ( Dwivedi et al., 2023 ). Different responses to prediction and simulation were observed between populations. As mentioned above, characteristics proper to each population limit direct comparisons and restrict results within each one. In a practical setting, each breeding program can evaluate the presence and inclusion of DelMut in the context of the already-implemented genotyping and selection approaches. Our results show consistent patterns of how the information on DelMut can benefit both GP and the design of crossing blocks. The MAGIC population had a lower PA and simulated genetic gains in yield compared to VEF and CDBN. Multiparent populations are designed to accumulate a greater number of recombination events, and the progeny lines are mosaics with contributions of all parents ( Gardner et al., 2016 ; Wang et al., 2022 ). These advantages make MAGIC populations particularly useful for gene mapping with a higher resolution, trait introgression, and marker-assisted selection. As more recombination events are accumulated, the linkage between alleles can be broken and allelic diversity increased. While this is desirable for eliminating DelMut in LD with beneficial alleles, this may also break the linkage between two or more positive alleles in coupling phase. In fact, Tourrette et al. (2019) and Taagen et al. (2022) showed that increased recombination, either in pericentromeric or chromosome-wide regions, can be detrimental for genetic gain and GP accuracies when it breaks up QTL in coupling phase. An increased recombination is particularly beneficial when positive QTL are in repulsion phase or targeted recombination approaches are adopted ( Ru and Bernardo, 2019 ). These may partly explain why the MAGIC population had lower PAs and simulated genetic gains, especially for yield where more QTL are expected to contribute to the trait variation (more complex). Additionally, Keller et al. (2020) observed lower PA in a different MAGIC population compared to the same VEF population used here. As proof of concept, the adopted approach has some limitations. In breeding programs, after each breeding cycle, new parents are integrated into the pipeline to preserve genetic diversity and continue increasing the genetic gain linearly. In our simulations, the genetic gain per cycle decreases after each cycle as alleles are fixed and the inbreeding increases. For the same reason, the cumulative genetic gain reaches a plateau value with no further increases. Here, we demonstrated in silico how the selection against DelMut has the potential to make breeding programs more efficient. Future simulations and empirical studies should evaluate simulations with the re-introduction of new germplasm similarly selected based on the number of DelMut. As a multifactor process, breeding programs are complex. Simulations that include changes in other parameters and evaluate the expected response of parents selected based on DelMut would also be valuable. Conclusions We evaluated whether predicting deleterious mutations in common bean breeding populations can enhance genomic prediction (GP) and increase the rate of genetic gain for yield and flowering time. Variants with predicted scores and with higher values had a highly leptokurtic gamma distribution, different from that of random and lower-scored markers. Incorporating variant effects and deleterious scores as weights in the GRM of a GBLUP model or in a multi-kernel approach for a BRR model improved yield prediction by 3.0 – 12.3%, while marginal improvements were observed for DTF (0.23 – 0.93%). The selection of parents with fewer highly DelMut has the potential of increasing the genetic gain in yield and flowering. For instance, in the VEF population, crosses between genotypes with seven to eight highly DelMut had a higher cumulative genetic gain for yield (5.85 Kg/Ha) compared to crosses with 12-13 highly DelMut (2.43 Kg/Ha). For flowering time, crossing schemes with fewer highly DelMut also showed more favorable DTF changes. These findings demonstrate the potential of DelMut information to support breeding program optimization through GP and simulation-based approaches. Data availability statement Code and materials used in this study can be found in the McGill University Pulse Breeding and Genetics GitHub page ( https://github.com/McGillHaricots/peas-andlove ) Acknowledgements We thank the Pulse Breeding and Genetics group members for their constructive feedback. We thank Isabella Chiaravalotti for the constructive discussions and guidance in the genomic prediction experiments, and Dr. Alice McQueen for facilitating the genotypic datasets for the CDBN population. References ↵ Alemu , A. , Åstrand , J. , Montesinos-López , O.A. , Isidro y Sánchez , J. , Fernández-Gónzalez , J. , Tadesse , W. , Vetukuri , R.R. , Carlsson , A.S. , Ceplitis , A. , Crossa , J. , Ortiz , R. , Chawade , A., 2024 . Genomic selection in plant breeding: Key factors shaping two decades of progress . Mol Plant 17 , 552 – 578 . doi: 10.1016/J.MOLP.2024.03.007 OpenUrl CrossRef PubMed ↵ Alley , E.C. , Khimulya , G. , Biswas , S. , AlQuraishi , M. , Church , G.M ., 2019 . Unified rational protein engineering with sequence-based deep representation learning . Nature Methods 2019 16:12 16 , 1315 – 1322 . doi: 10.1038/s41592-019-0598-1 OpenUrl CrossRef PubMed ↵ Amadeu , R.R. , Garcia , A.A.F. , Munoz , P.R. , Ferrão , L.F. V ., 2023 . AGHmatrix: genetic relationship matrices in R . Bioinformatics 39 . doi: 10.1093/BIOINFORMATICS/BTAD445 OpenUrl CrossRef ↵ Ates , D. , Asciogul , T.K. , Nemli , S. , Erdogmus , S. , Esiyok , D. , Tanyolac , M.B ., 2018 . Association mapping of days to flowering in common bean (Phaseolus vulgaris L.) revealed by DArT markers . Molecular Breeding 38 . doi: 10.1007/s11032-018-0868-0 OpenUrl CrossRef ↵ Bates , D. , Mächler , M. , Bolker , B.M. , Walker , S.C ., 2015 . Fitting linear mixed-effects models using lme4 . J Stat Softw 67 . doi: 10.18637/jss.v067.i01 OpenUrl CrossRef PubMed ↵ Böndel , K.B. , Samuels , T. , Craig , R.J. , Ness , R.W. , Colegrave , N. , Keightley , P.D ., 2022 . The distribution of fitness effects of spontaneous mutations in Chlamydomonas reinhardtii inferred using frequency changes under experimental evolution . PLoS Genet 18 , e1009840 . doi: 10.1371/JOURNAL.PGEN.1009840 OpenUrl CrossRef PubMed ↵ Bradbury , P.J. , Casstevens , T. , Jensen , S.E. , Johnson , L.C. , Miller , Z.R. , Monier , B. , Romay , M.C. , Song , B. , Buckler , E.S ., 2022 . The Practical Haplotype Graph, a platform for storing and using pangenomes for imputation . Bioinformatics 38 , 3698 – 3702 . doi: 10.1093/BIOINFORMATICS/BTAC410 OpenUrl CrossRef PubMed ↵ Bradbury , P.J. , Zhang , Z. , Kroon , D.E. , Casstevens , T.M. , Ramdoss , Y. , Buckler , E.S ., 2007 . TASSEL: Software for association mapping of complex traits in diverse samples . Bioinformatics 23 , 2633 – 2635 . doi: 10.1093/bioinformatics/btm308 OpenUrl CrossRef PubMed Web of Science ↵ Chang , C.C. , Chow , C.C. , Tellier , L.C.A.M. , Vattikuti , S. , Purcell , S.M. , Lee , J.J ., 2015 . Second-generation PLINK: Rising to the challenge of larger and richer datasets . Gigascience 4 , 7 . doi: 10.1186/S13742-015-0047-8/2707533 OpenUrl CrossRef ↵ Charlesworth , B. , Charlesworth , D ., 1999 . The genetic basis of inbreeding depression . Genet Res (Camb ) 74 , 329 – 340 . doi: 10.1017/S0016672399004152 OpenUrl CrossRef PubMed Web of Science ↵ Charlesworth , B. , Morgan , M.T. , Charlesworth , D ., 1993 . The effect of deleterious mutations on neutral molecular variation . Genetics 134 , 1289 – 1303 . doi: 10.1093/GENETICS/134.4.1289 OpenUrl Abstract / FREE Full Text ↵ Chiaravallotti , I. , Lin , J. , Arief , V. , Jahufer , Z. , Osorno , J.M. , McClean , P. , Jarquin , D. , Hoyos-Villegas , V ., 2024 . Simulations of multiple breeding strategy scenarios in common bean for assessing genomic selection accuracy and model updating . Plant Genome 17 , e20388 . doi: 10.1002/TPG2.20388 OpenUrl CrossRef ↵ Chris Gaynor , R. , Gorjanc , G. , Hickey , J.M. , 2021 . AlphaSimR: an R package for breeding program simulations . G3 Genes|Genomes|Genetics 11 . doi: 10.1093/G3JOURNAL/JKAA017 OpenUrl CrossRef ↵ Covarrubias-Pazaran , G. , Gebeyehu , Z. , Gemenet , D. , Werner , C. , Labroo , M. , Sirak , S. , Coaldrake , P. , Rabbi , I. , Kayondo , S.I. , Parkes , E. , Kanju , E. , Mbanjo , E.G.N. , Agbona , A. , Kulakow , P. , Quinn , M. , Debaene , J ., 2022 . Breeding Schemes: What Are They, How to Formalize Them, and How to Improve Them? Front Plant Sci 12 , 791859 . doi: 10.3389/FPLS.2021.791859/BIBTEX OpenUrl CrossRef PubMed ↵ Cordoba-Novoa , H. Buckler , E.S ; Romay , C.M. ; Berthel , A ; Johnson , L ; Balasubramanian , P ; Hoyos-Villegas , V. 2025 . Phylogenetic Analysis and Machine Learning Identify Signatures of Selection and Predict Deleterious Mutations in Common Bean (Phaseolus vulgaris L.) . Thesis . McGill University . Montreal, Canada . ↵ Davydov , E. V. , Goode , D.L. , Sirota , M. , Cooper , G.M. , Sidow , A. , Batzoglou , S ., 2010 . Identifying a High Fraction of the Human Genome to be under Selective Constraint Using GERP++ . PLoS Comput Biol 6 , e1001025 . doi: 10.1371/JOURNAL.PCBI.1001025 OpenUrl CrossRef PubMed ↵ de los Campos , G. , Hickey , J.M. , Pong-Wong , R. , Daetwyler , H.D. , Calus , M.P.L. , 2013 . Whole-Genome Regression and Prediction Methods Applied to Plant and Animal Breeding . Genetics 193 , 327 – 345 . doi: 10.1534/GENETICS.112.143313 OpenUrl Abstract / FREE Full Text ↵ Delfini , J. , Moda-Cirino , V. , dos Santos Neto , J. , Zeffa , D.M. , Nogueira , A.F. , Ribeiro , L.A.B. , Ruas , P.M. , Gepts , P. , Gonçalves , L.S.A. , 2021 . Genome-Wide Association Study Identifies Genomic Regions for Important Morpho-Agronomic Traits in Mesoamerican Common Bean . Front Plant Sci 12 , 748829 . doi: 10.3389/FPLS.2021.748829/BIBTEX OpenUrl CrossRef PubMed ↵ Diaz , S. , Ariza-Suarez , D. , Izquierdo , P. , Lobaton , J.D. , de la Hoz , J.F. , Acevedo , F. , Duitama , J. , Guerrero , A.F. , Cajiao , C. , Mayor , V. , Beebe , S.E. , Raatz , B. , 2020 . Genetic mapping for agronomic traits in a MAGIC population of common bean (Phaseolus vulgaris L.) under drought conditions . BMC Genomics 21 . doi: 10.1186/s12864-020-07213-6 OpenUrl CrossRef PubMed ↵ Dussex , N. , Morales , H.E. , Grossen , C. , Dalén , L. , van Oosterhout , C. , 2023 . Purging and accumulation of genetic load in conservation . Trends Ecol Evol 38 , 961 – 969 . doi: 10.1016/J.TREE.2023.05.008 OpenUrl CrossRef PubMed ↵ Dwivedi , S.L. , Heslop-Harrison , P. , Spillane , C. , McKeown , P.C. , Edwards , D. , Goldman , I. , Ortiz , R ., 2023 . Evolutionary dynamics and adaptive benefits of deleterious mutations in crop gene pools . Trends Plant Sci 28 , 685 – 697 . doi: 10.1016/J.TPLANTS.2023.01.006/ASSET/5671AB81-1B75-4D93-BE10-5440E83207E8/MAIN.ASSETS/GR1.JPG OpenUrl CrossRef PubMed ↵ Edwards , S.M. , Sørensen , I.F. , Sarup , P. , Mackay , T.F.C. , Sørensen , P ., 2016 . Genomic prediction for quantitative traits is improved by mapping variants to gene ontology categories in Drosophila melanogaster . Genetics 203 , 1871 – 1883 . doi: 10.1534/GENETICS.116.187161/-/DC1 OpenUrl Abstract / FREE Full Text ↵ Endelman , J.B ., 2011 . Ridge Regression and Other Kernels for Genomic Selection with R Package rrBLUP . Plant Genome 4 , 250 – 255 . doi: 10.3835/PLANTGENOME2011.08.0024 OpenUrl CrossRef ↵ Eyre-Walker , A ., 2006 . The genomic rate of adaptive evolution . Trends Ecol Evol 21 , 569 – 575 . doi: 10.1016/j.tree.2006.06.015 OpenUrl CrossRef PubMed Web of Science ↵ Eyre-Walker , A. , Keightley , P.D ., 2007 . The distribution of fitness effects of new mutations . Nature Reviews Genetics 2007 8:8 8 , 610 – 618 . doi: 10.1038/nrg2146 OpenUrl CrossRef PubMed Web of Science ↵ Eyre-Walker , A. , Woolfit , M. , Phelps , T ., 2006 . The Distribution of Fitness Effects of New Deleterious Amino Acid Mutations in Humans . Genetics 173 , 891 – 900 . doi: 10.1534/GENETICS.106.057570 OpenUrl Abstract / FREE Full Text ↵ Fang , L. , Sahana , G. , Ma , P. , Su , G. , Yu , Y. , Zhang , S. , Lund , M.S. , Sørensen , P ., 2017 . Exploring the genetic architecture and improving genomic prediction accuracy for mastitis and milk production traits in dairy cattle by mapping variants to hepatic transcriptomic regions responsive to intra-mammary infection . Genetics Selection Evolution 49 , 1 – 18 . doi: 10.1186/S12711-017-0319-0/FIGURES/6 OpenUrl CrossRef PubMed ↵ Felsenstein , J ., 1974 . The Evolutionary Advantage of Recombination . Genetics 78 , 737 – 756 . doi: 10.1093/GENETICS/78.2.737 OpenUrl Abstract / FREE Full Text ↵ Fritsche-Neto , R. , Ali , J. , De Asis , E.J. , Allahgholipour , M. , Labroo , M.R. , 2024 . Improving hybrid rice breeding programs via stochastic simulations: number of parents, number of hybrids, tester update, and genomic prediction of hybrid performance . Theoretical and Applied Genetics 137 , 1 – 12 . doi: 10.1007/S00122-023-04508-6/FIGURES/5 OpenUrl CrossRef ↵ Gao , C ., 2021 . Genome engineering for crop improvement and future agriculture . Cell 184 , 1621 – 1635 . doi: 10.1016/J.CELL.2021.01.005 OpenUrl CrossRef PubMed ↵ Gardner , K.A. , Wittern , L.M. , Mackay , I.J ., 2016 . A highly recombined, high-density, eight-founder wheat MAGIC map reveals extensive segregation distortion and genomic locations of introgression segments . Plant Biotechnol J 14 , 1406 – 1417 . doi: 10.1111/PBI.12504 OpenUrl CrossRef PubMed ↵ Glaus , A.N. , Brechet , M. , Swinnen , G. , Lebeigle , L. , Iwaszkiewicz , J. , Ambrosini , G. , Julca , I. , Zhang , J. , Roberts , R. , Iseli , C. , Guex , N. , Jiménez-Gómez , J. , Glover , N. , Martin , G.B. , Strickler , S. , Soyk , S ., 2025 . Repairing a deleterious domestication variant in a floral regulator gene of tomato by base editing . Nature Genetics 2025 57:1 57 , 231 – 241 . doi: 10.1038/s41588-024-02026-9 OpenUrl CrossRef PubMed ↵ Glémin , S ., 2003 . HOW ARE DELETERIOUS MUTATIONS PURGED? DRIFT VERSUS NONRANDOM MATING . Evolution (N Y ) 57 , 2678 – 2687 . doi: 10.1111/J.0014-3820.2003.TB01512.X OpenUrl CrossRef ↵ Goddard , M.E. , Hayes , B.J. , Meuwissen , T.H.E ., 2010 . Genomic selection in livestock populations . Genet Res (Camb ) 92 , 413 – 421 . doi: 10.1017/S0016672310000613 OpenUrl CrossRef PubMed ↵ Gorjanc , G. , Gaynor , R.C. , Hickey , J.M ., 2018 . Optimal cross selection for long-term genetic gain in two-part programs with rapid recurrent genomic selection . Theoretical and Applied Genetics 131 , 1953 – 1966 . doi: 10.1007/S00122-018-3125-3/FIGURES/6 OpenUrl CrossRef PubMed ↵ Gossmann , T.I. , Keightley , P.D. , Eyre-Walker , A ., 2012 . The effect of variation in the effective population size on the rate of adaptive molecular evolution in eukaryotes . Genome Biol Evol 4 , 658 – 667 . doi: 10.1093/GBE/EVS027 OpenUrl CrossRef PubMed ↵ Grossen , C. , Ramakrishnan , U ., 2024 . Genetic load . Current Biology 34 , R1216 – R1220 . doi: 10.1016/j.cub.2024.11.004 OpenUrl CrossRef PubMed ↵ Hassanpour , A. , Geibel , J. , Simianer , H. , Pook , T ., 2023 . Optimization of breeding program design through stochastic simulation with kernel regression . G3 Genes|Genomes|Genetics 13 . doi: 10.1093/G3JOURNAL/JKAD217 OpenUrl CrossRef ↵ Johnsson , M. , Gaynor , R.C. , Jenko , J. , Gorjanc , G. , De Koning , D.J. , Hickey , J.M. , 2019 . Removal of alleles by genome editing (RAGE) against deleterious load . Genetics Selection Evolution 51 , 1 – 18 . doi: 10.1186/S12711-019-0456-8/FIGURES/10 OpenUrl CrossRef PubMed ↵ Kamfwa , K. , Cichy , K.A. , Kelly , J.D ., 2015 . Genome-Wide Association Study of Agronomic Traits in Common Bean . Plant Genome 8 . doi: 10.3835/plantgenome2014.09.0059 OpenUrl CrossRef ↵ Keightley , P.D ., 1996 . Nature of Deleterious Mutation Load in Drosophila . Genetics 144 , 1993 – 1999 . doi: 10.1093/GENETICS/144.4.1993 OpenUrl Abstract / FREE Full Text ↵ Keller , B. , Ariza-Suarez , D. , de la Hoz , J. , Aparicio , J.S. , Portilla-Benavides , A.E. , Buendia , H.F. , Mayor , V.M. , Studer , B. , Raatz , B. , 2020 . Genomic Prediction of Agronomic Traits in Common Bean (Phaseolus vulgaris L.) Under Environmental Stress . Front Plant Sci 11 , 543352 . doi: 10.3389/FPLS.2020.01001/BIBTEX OpenUrl CrossRef ↵ Kim , M.S. , Lozano , R. , Kim , J.H. , Bae , D.N. , Kim , S.T. , Park , J.H. , Choi , M.S. , Kim , J. , Ok , H.C. , Park , S.K. , Gore , M.A. , Moon , J.K. , Jeong , S.C ., 2021 . The patterns of deleterious mutations during the domestication of soybean . Nat Commun 12 , 1 – 14 . doi: 10.1038/s41467-020-20337-3 OpenUrl CrossRef ↵ Klápště , J. , Dungey , H.S. , Telfer , E.J. , Suontama , M. , Graham , N.J. , Li , Y. , McKinley , R ., 2020 . Marker Selection in Multivariate Genomic Prediction Improves Accuracy of Low Heritability Traits . Front Genet 11 , 499094 . doi: 10.3389/FGENE.2020.499094/BIBTEX OpenUrl CrossRef PubMed ↵ Kono , T.J.Y. , Fu , F. , Mohammadi , M. , Hoffman , P.J. , Liu , C. , Stupar , R.M. , Smith , K.P. , Tiffin , P. , Fay , J.C. , Morrell , P.L ., 2016 . The Role of Deleterious Substitutions in Crop Genomes . Mol Biol Evol 33 , 2307 – 2317 . doi: 10.1093/molbev/msw102 OpenUrl CrossRef PubMed ↵ Kono , T.J.Y. , Liu , C. , Vonderharr , E.E. , Koenig , D. , Fay , J.C. , Smith , K.P. , Morrell , P.L ., 2019 . The Fate of Deleterious Variants in a Barley Genomic Prediction Population . Genetics 213 , 1531 – 1544 . doi: 10.1534/GENETICS.119.302733 OpenUrl Abstract / FREE Full Text ↵ Krasovec , M. , Eyre-Walker , A. , Grimsley , N. , Salmeron , C. , Pecqueur , D. , Piganeau , G. , Sanchez-Ferandin , S ., 2016 . Fitness effects of spontaneous mutations in picoeukaryotic marine green algae . G3: Genes, Genomes, Genetics 6 , 2063 – 2071 . doi: 10.1534/G3.116.029769/-/DC1 OpenUrl CrossRef ↵ Lin , J. , Arief , V. , Jahufer , Z. , Osorno , J. , McClean , P. , Jarquin , D. , Hoyos-Villegas , V ., 2023 . Simulations of rate of genetic gain in dry bean breeding programs . Theoretical and Applied Genetics 136 , 1 – 22 . doi: 10.1007/S00122-023-04244-X/FIGURES/11 OpenUrl CrossRef PubMed ↵ Liu , A. , Lund , M.S. , Boichard , D. , Karaman , E. , Guldbrandtsen , B. , Fritz , S. , Aamand , G.P. , Nielsen , U.S. , Sahana , G. , Wang , Y. , Su , G ., 2020 . Weighted single-step genomic best linear unbiased prediction integrating variants selected from sequencing data by association and bioinformatics analyses . Genetics Selection Evolution 52 , 1 – 17 . doi: 10.1186/S12711-020-00568-0/TABLES/6 OpenUrl CrossRef PubMed ↵ Liu , H. , Tessema , B.B. , Jensen , J. , Cericola , F. , Andersen , J.R. , Sørensen , A.C ., 2019 . ADAM-Plant: A software for stochastic simulations of plant breeding from molecular to phenotypic level and from simple selection to complex speed breeding programs . Front Plant Sci 9 , 425945 . doi: 10.3389/FPLS.2018.01926/BIBTEX OpenUrl CrossRef ↵ Long , E.M. , Romay , M.C. , Ramstein , G. , Buckler , E.S. , Robbins , K.R ., 2023 . Utilizing evolutionary conservation to detect deleterious mutations and improve genomic prediction in cassava . Front Plant Sci 13 . doi: 10.3389/fpls.2022.1041925 OpenUrl CrossRef ↵ MacQueen , A.H. , White , J.W. , Lee , R. , Osorno , J.M. , Schmutz , J. , Miklas , P.N. , Myers , J. , McClean , P.E. , Juenger , T.E ., 2020 . Genetic Associations in Four Decades of Multi-Environment Trials Reveal Agronomic Trait Evolution in Common Bean . Genetics 215 , 267 – 284 . doi: 10.1101/734087 OpenUrl Abstract / FREE Full Text ↵ Meuwissen , T. , Eikje , L.S. , Gjuvsland , A.B ., 2024 . GWABLUP: genome-wide association assisted best linear unbiased prediction of genetic values . Genetics Selection Evolution 56 , 1 – 12 . doi: 10.1186/S12711-024-00881-Y/TABLES/5 OpenUrl CrossRef PubMed ↵ Moghaddam , S.M. , Mamidi , S. , Osorno , J.M. , Lee , R. , Brick , M. , Kelly , J. , Miklas , P. , Urrea , C. , Song , Q. , Cregan , P. , Grimwood , J. , Schmutz , J. , McClean , P.E ., 2016 . Genome-Wide Association Study Identifies Candidate Loci Underlying Agronomic Traits in a Middle American Diversity Panel of Common Bean . Plant Genome 9 , 1 – 21 . doi: 10.3835/plantgenome2016.02.0012 OpenUrl CrossRef ↵ Morgante , F. , Huang , W. , Maltecca , C. , Mackay , T.F.C ., 2018 . Effect of genetic architecture on the prediction accuracy of quantitative traits in samples of unrelated individuals . Heredity 2018 120:6 120 , 500 – 514 . doi: 10.1038/s41437-017-0043-0 OpenUrl CrossRef PubMed ↵ Moyers , B.T. , Morrell , P.L. , McKay , J.K ., 2018 . Genetic costs of domestication and improvement . Journal of Heredity 109 , 103 – 116 . doi: 10.1093/jhered/esx069 OpenUrl CrossRef PubMed ↵ Nascimento , M. , Nascimento , A.C.C. , Silva , F.F. e. , Barili , L.D. , Do Vale , N.M. , Carneiro , J.E. , Cruz , C.D. , Carneiro , P.C.S. , Serão , N.V.L ., 2018 . Quantile regression for genome-wide association study of flowering time-related traits in common bean . PLoS One 13 , 1 – 14 . doi: 10.1371/journal.pone.0190303 OpenUrl CrossRef PubMed ↵ Ng , P.C. , Henikoff , S ., 2003 . SIFT: predicting amino acid changes that affect protein function . Nucleic Acids Res 31 , 3812 – 3814 . doi: 10.1093/NAR/GKG509 OpenUrl CrossRef PubMed Web of Science ↵ Ohta , T. , Kimura , M ., 1971 . On the constancy of the evolutionary rate of cistrons . J Mol Evol 1 , 18 – 25 . doi: 10.1007/BF01659391/METRICS OpenUrl CrossRef PubMed ↵ Orr , H.A ., 2009 . Fitness and its role in evolutionary genetics . Nature Reviews Genetics 2009 10:8 10 , 531 – 539 . doi: 10.1038/nrg2603 OpenUrl CrossRef PubMed Web of Science ↵ Peixoto , M.A. , Coelho , I.F. , Leach , K.A. , Lübberstedt , T. , Bhering , L.L. , Resende , M.F.R ., 2024 . Use of simulation to optimize a sweet corn breeding program: implementing genomic selection and doubled haploid technology . G3 Genes|Genomes|Genetics 14 . doi: 10.1093/G3JOURNAL/JKAE128 OpenUrl CrossRef ↵ Pérez , P. , De Los Campos , G. , 2014 . Genome-wide regression and prediction with the BGLR statistical package . Genetics 198 , 483 – 495 . doi: 10.1534/GENETICS.114.164442/-/DC1 OpenUrl Abstract / FREE Full Text ↵ Piganeau , G. , Eyre-Walker , A ., 2003 . Estimating the distribution of fitness effects from DNA sequence data: Implications for the molecular clock . Proc Natl Acad Sci U S A 100 , 10335 – 10340 . doi: 10.1073/PNAS.1833064100/SUPPL_FILE/3064TABLE5.HTML OpenUrl Abstract / FREE Full Text ↵ Raggi , L. , Caproni , L. , Carboni , A. , Negri , V ., 2019 . Genome-Wide Association Study Reveals Candidate Genes for Flowering Time Variation in Common Bean (Phaseolus vulgaris L .). Front Plant Sci 10 , 1 – 14 . doi: 10.3389/fpls.2019.00962 OpenUrl CrossRef PubMed ↵ Ramstein , G.P. , Buckler , E.S ., 2022 . Prediction of evolutionary constraint by genomic annotations improves functional prioritization of genomic variants in maize . Genome Biol 23 . doi: 10.1186/s13059-022-02747-2 OpenUrl CrossRef ↵ Raoul , J. , Palhière , I. , Astruc , J.M. , Swan , A. , Elsen , J.M ., 2018 . Optimal mating strategies to manage a heterozygous advantage major gene in sheep . Animal 12 , 454 – 463 . doi: 10.1017/S1751731117001835 OpenUrl CrossRef PubMed ↵ Renaut , S. , Rieseberg , L.H ., 2015 . The Accumulation of Deleterious Mutations as a Consequence of Domestication and Improvement in Sunflowers and Other Compositae Crops . Mol Biol Evol 32 , 2273 – 2283 . doi: 10.1093/MOLBEV/MSV106 OpenUrl CrossRef PubMed ↵ Robinson , J. , Kyriazis , C.C. , Yuan , S.C. , Lohmueller , K.E ., 2023 . Deleterious Variation in Natural Populations and Implications for Conservation Genetics . Annu Rev Anim Biosci 11 , 93 – 114 . doi: 10.1146/ANNUREV-ANIMAL-080522-093311/CITE/REFWORKS OpenUrl CrossRef PubMed ↵ Ru , S. , Bernardo , R ., 2019 . Targeted recombination to increase genetic gain in self-pollinated species . Theoretical and Applied Genetics 132 , 289 – 300 . doi: 10.1007/S00122-018-3216-1/FIGURES/3 OpenUrl CrossRef PubMed ↵ Shin , J.H. , Blay , S. , McNeney , B. , Graham , J ., 2006 . LDheatmap: An R Function for Graphical Display of Pairwise Linkage Disequilibria Between Single Nucleotide Polymorphisms . J Stat Softw 16 , 1 – 9 . doi: 10.18637/JSS.V016.C03 OpenUrl CrossRef ↵ Silva , É.D.B. da , Xavier , A. , Faria , M.V. , 2021 . Impact of Genomic Prediction Model, Selection Intensity, and Breeding Strategy on the Long-Term Genetic Gain and Genetic Erosion in Soybean Breeding . Front Genet 12 , 637133 . doi: 10.3389/FGENE.2021.637133/ENDNOTE OpenUrl CrossRef PubMed ↵ Sonesson , A.K. , Janss , L.L.G. , Meuwissen , T.H.E ., 2003 . Selection against genetic defects in conservation schemes while controlling inbreeding . Genetics Selection Evolution 2003 35:5 35 , 1 – 16 . doi: 10.1186/1297-9686-35-5-353 OpenUrl CrossRef ↵ Sun , S. , Wang , B. , Li , C. , Xu , G. , Yang , J. , Hufford , M.B. , Ross-Ibarra , J. , Wang , H. , Wang , L ., 2023 . Unraveling Prevalence and Effects of Deleterious Mutations in Maize Elite Lines across Decades of Modern Breeding . Mol Biol Evol 40 . doi: 10.1093/MOLBEV/MSAD170 OpenUrl CrossRef ↵ Taagen , E. , Jordan , K. , Akhunov , E. , Sorrells , M.E. , Jannink , J.L ., 2022 . If it ain’t broke, don’t fix it: evaluating the effect of increased recombination on response to selection for wheat breeding . G3 Genes|Genomes|Genetics 12 . doi: 10.1093/G3JOURNAL/JKAC291 OpenUrl CrossRef ↵ Tourrette , E. , Bernardo , R. , Falque , M. , Martin , O.C ., 2019 . Assessing by Modeling the Consequences of Increased Recombination in Recurrent Selection of Oryza sativa and Brassica rapa . G3 Genes|Genomes|Genetics 9 , 4169 – 4181 . doi: 10.1534/G3.119.400545 OpenUrl CrossRef ↵ Valluru , R. , Gazave , E.E. , Fernandes , S.B. , Ferguson , J.N. , Lozano , R. , Hirannaiah , P. , Zuo , T. , Brown , P.J. , Leakey , A.D.B. , Gore , M.A. , Buckler , E.S. , Bandillo , N ., 2019 . Deleterious mutation burden and its association with complex traits in sorghum (Sorghum bicolor) . Genetics 211 , 1075 – 1087 . doi: 10.1534/genetics.118.301742 OpenUrl Abstract / FREE Full Text ↵ VanRaden , P.M ., 2008 . Efficient methods to compute genomic predictions . J Dairy Sci 91 , 4414 – 4423 . doi: 10.3168/jds.2007-0980 OpenUrl CrossRef PubMed Web of Science ↵ VanRaden , P.M. , Tooker , M.E. , O’Connell , J.R. , Cole , J.B. , Bickhart , D.M ., 2017 . Selecting sequence variants to improve genomic predictions for dairy cattle . Genetics Selection Evolution 49 , 1 – 12 . doi: 10.1186/S12711-017-0307-4/FIGURES/4 OpenUrl CrossRef PubMed ↵ Vieira , R.A. , Nogueira , A.P.O. , Fritsche-Neto , R ., 2025 . Optimizing the selection of quantitative traits in plant breeding using simulation . Front Plant Sci 16 , 1495662 . doi: 10.3389/FPLS.2025.1495662/BIBTEX OpenUrl CrossRef PubMed ↵ Vilhjálmsson , B.J. , Nordborg , M ., 2012 . The nature of confounding in genome-wide association studies . Nature Reviews Genetics 2012 14:1 14 , 1 – 2 . doi: 10.1038/nrg3382 OpenUrl CrossRef PubMed ↵ Wallace , J.G. , Rodgers-Melnick , E. , Buckler , E.S ., 2018 . On the road to breeding 4.0: Unraveling the good, the bad, and the boring of crop quantitative genomics . Annu Rev Genet 52 , 421 – 444 . doi: 10.1146/ANNUREV-GENET-120116-024846/CITE/REFWORKS OpenUrl CrossRef PubMed ↵ Wang , J. , Zhang , Z ., 2021 . GAPIT Version 3: Boosting Power and Accuracy for Genomic Association and Prediction . Genomics Proteomics Bioinformatics 19 , 629 – 640 . doi: 10.1016/J.GPB.2021.08.005 OpenUrl CrossRef ↵ Wang , J. , Zhou , Z. , Zhang , Zhe , Li , H. , Liu , D. , Zhang , Q. , Bradbury , P.J. , Buckler , E.S. , Zhang , Zhiwu , 2018 . Expanding the BLUP alphabet for genomic prediction adaptable to the genetic architectures of complex traits . Heredity 2018 121:6 121 , 648 – 662 . doi: 10.1038/s41437-018-0075-0 OpenUrl CrossRef PubMed ↵ Wang , M. , Qi , Z. , Thyssen , G.N. , Naoumkina , M. , Jenkins , J.N. , McCarty , J.C. , Xiao , Y. , Li , J. , Zhang , X. , Fang , D.D ., 2022 . Genomic interrogation of a MAGIC population highlights genetic factors controlling fiber quality traits in cotton . Communications Biology 2022 5:1 5 , 1 – 12 . doi: 10.1038/s42003-022-03022-7 OpenUrl CrossRef PubMed ↵ Wu , Y. , Li , D. , Hu , Y. , Li , H. , Ramstein , G.P. , Zhou , S. , Zhang , X. , Bao , Z. , Zhang , Y. , Song , B. , Zhou , Yao , Zhou , Yongfeng , Gagnon , E. , Särkinen , T. , Knapp , S. , Zhang , C. , Städler , T. , Buckler , E.S. , Huang , S ., 2023 . Phylogenomic discovery of deleterious mutations facilitates hybrid potato breeding . Cell 186 , 2313 – 2328 .e15. doi: 10.1016/J.CELL.2023.04.008 OpenUrl CrossRef PubMed ↵ Xavier , A ., 2019 . Efficient Estimation of Marker Effects in Plant Breeding . G3 Genes|Genomes|Genetics 9 , 3855 – 3866 . doi: 10.1534/G3.119.400728 OpenUrl CrossRef ↵ Yang , J. , Mezmouk , S. , Baumgarten , A. , Buckler , E.S. , Guill , K.E. , McMullen , M.D. , Mumm , R.H. , Ross-Ibarra , J ., 2017 . Incomplete dominance of deleterious alleles contributes substantially to trait variation and heterosis in maize . PLoS Genet 13 . doi: 10.1371/journal.pgen.1007019 OpenUrl CrossRef PubMed ↵ Zhang , Y. , Zhuang , Z. , Liu , Y. , Huang , J. , Luan , M. , Zhao , X. , Dong , L. , Ye , J. , Yang , M. , Zheng , E. , Cai , G. , Wu , Z. , Yang , J ., 2024 . Genomic prediction based on preselected single-nucleotide polymorphisms from genome-wide association study and imputed whole-genome sequence data annotation for growth traits in Duroc pigs . Evol Appl 17 , e13651 . doi: 10.1111/EVA.13651 OpenUrl CrossRef PubMed ↵ Zhu , M. , Cheng , Y. , Wu , S. , Huang , X. , Qiu , J ., 2022 . Deleterious mutations are characterized by higher genomic heterozygosity than other genic variants in plant genomes . Genomics 114 , 110290 . doi: 10.1016/J.YGENO.2022.110290 OpenUrl CrossRef PubMed View the discussion thread. Back to top Previous Next Posted May 09, 2025. Download PDF Email Thank you for your interest in spreading the word about bioRxiv. NOTE: Your email address is requested solely to identify you as the sender of this article. Your Email * Your Name * Send To * Enter multiple addresses on separate lines or separate them with commas. You are going to email the following Prioritization of Deleterious Mutations Improves Genomic Prediction and Increases the Rate of Genetic Gain in Common Bean (Phaseolus vulgaris L.), a Simulation Study Message Subject (Your Name) has forwarded a page to you from bioRxiv Message Body (Your Name) thought you would like to see this page from the bioRxiv website. Your Personal Message CAPTCHA This question is for testing whether or not you are a human visitor and to prevent automated spam submissions. Share Prioritization of Deleterious Mutations Improves Genomic Prediction and Increases the Rate of Genetic Gain in Common Bean ( Phaseolus vulgaris L. ), a Simulation Study H. Cordoba-Novoa , V. Hoyos-Villegas bioRxiv 2025.05.05.652208; doi: https://doi.org/10.1101/2025.05.05.652208 Share This Article: Copy Citation Tools Prioritization of Deleterious Mutations Improves Genomic Prediction and Increases the Rate of Genetic Gain in Common Bean ( Phaseolus vulgaris L. ), a Simulation Study H. Cordoba-Novoa , V. Hoyos-Villegas bioRxiv 2025.05.05.652208; doi: https://doi.org/10.1101/2025.05.05.652208 Citation Manager Formats BibTeX Bookends EasyBib EndNote (tagged) EndNote 8 (xml) Medlars Mendeley Papers RefWorks Tagged Ref Manager RIS Zotero Tweet Widget Facebook Like Google Plus One Subject Area Genetics Subject Areas All Articles Animal Behavior and Cognition (7618) Biochemistry (17635) Bioengineering (13859) Bioinformatics (41846) Biophysics (21401) Cancer Biology (18534) Cell Biology (25422) Clinical Trials (138) Developmental Biology (13352) Ecology (19860) Epidemiology (2067) Evolutionary Biology (24285) Genetics (15582) Genomics (22463) Immunology (17700) Microbiology (40298) Molecular Biology (17141) Neuroscience (88424) Paleontology (666) Pathology (2825) Pharmacology and Toxicology (4813) Physiology (7633) Plant Biology (15107) Scientific Communication and Education (2042) Synthetic Biology (4284) Systems Biology (9808) Zoology (2267)