Identification of regulatory loci and candidate genes related to body weight traits in broilers based on different models

BACKGROUND

Growth traits are crucial for the economic viability in broiler production, as they significantly contribute to the cost of rearing. Maximizing body weight (BW) while minimizing feed intake is key to enhancing the efficiency of broiler breeding. Identifying the genetic architecture associated with BW trait is therefore a critical step in enhancing breeding strategies.

RESULTS

We conducted a genome-wide association study (GWAS) using two statistical approaches: single-trait GWAS and longitudinal GWAS. The study was performed on the BW trait at five developmental stages (72, 81, 89, 113, and 120 days) and mid-test metabolic weight (MWT) across four growth cycles. Transcriptome sequencing analysis was also included to investigate the differential expression of candidate genes identified through the GWAS models, particularly linked to BW and MWT traits. Utilizing the chicken 55K single nucleotide polymorphism (SNP) array, we identified 52,060 SNPs in the genomic data of 4,493 Wenchang chickens. The single-trait GWAS model revealed 42 BW-associated SNPs, corresponding to 18 potential genes. For MWT, 47 SNPs were associated, mapping to 31 candidate genes. The longitudinal GWAS model identified 34 BW-linked SNPs, annotated with 22 candidate genes, and 21 MWT-linked SNPs, annotated with 10 candidate genes. Notably, 16 SNPs on chromosome 4 were associated with both BW and MWT, located within the 73.08Mb-76.82Mb region. Nine genes were annotated from this region, including STIM2, SEL1L3, SEPSECS, LGI2, SOD3, KCNIP4, NCAPG, FAM184B, LDB2. Notably, there are 32 overlapping SNPs identified in both the single-trait and longitudinal GWAS models, suggesting consistent associations for both BW and MWT. These overlapping SNPs represent robust loci that may influence both traits across different statistical approaches. Transcriptome sequencing indicated differential expression of LDB2 and SEL1L3 between high and low BW groups.

CONCLUSION

Our study has uncovered novel candidate genes that are potentially involved in growth traits, providing valuable insights for broiler breeding. The identified SNPs and genes could serve as genetic markers for selecting broilers with improved growth efficiency, which may lead to more cost-effective and productive broiler farming.

Genome-Wide Association Study of Body Size Traits in Luning Chickens Using Whole-Genome Sequencing

Growth traits are crucial for poultry breeding and production. Marker-assisted selection (MAS) and genomic selection (GS) of growth traits require a substantial number of accurate genetic markers. A genome-wide association study (GWAS) for body size traits was performed on 248 Luning chickens to identify significant single-nucleotide polymorphisms (SNPs) and insertions and deletions (INDELs) related to the growth and development of chickens. A total of 30 significant SNPs and 13 INDELs were obtained for body size traits. Two notable regions, spanning from 43.072 to 43.219 Mb on chromosome 1 and from 4.751 to 4.800 Mb on chromosome 11, were found to be significantly associated with growth traits in the GWAS of both SNPs and INDELs. Some genes, including PPFIA2, KITLG, DUSP6, TOX3, MTNR1B, FAT3, PTPRR, VEZT, BBS9, and CYLD, were identified as important candidate genes for the growth of chickens. The results provide valuable information for understanding the genetic basis of growth traits which is beneficial for the subsequent selective breeding in Luning chickens.

Genome-Wide Association Study of Insertions and Deletions Identified Novel Loci Associated with Milk Production Traits in Dairy Cattle

Genome-wide association study (GWAS) have identified a large number of SNPs associated with milk production traits in dairy cattle. Behind SNPs, INDELs are the second most abundant genetic polymorphisms in the genome, which may exhibit an independent association with complex traits in humans and other species. However, there are no reports on GWASs of INDELs for milk production traits in dairy cattle. In this study, using imputed sequence data, we performed INDEL-based and SNP-based GWASs for milk production traits in a Holstein cattle population. We identified 58 unique significant INDELs for one or multiple traits. The majority of these INDELs are in considerable LD with nearby significant SNPs. However, through conditional association analysis, we identified nine INDELs which showed independent associations. Genomic annotations of these INDELs indicated some novel associated genes, i.e., TRNAG-CCC, EPPK1, PPM1K, PTDSS1, and mir-10163, which were not reported in previous SNP-based GWASs. Our findings suggest that INDEL-based GWASs could be valuable complement to SNP-based GWASs for milk production traits.

Whole genome sequencing reveals that five genes are related to BW trait in sheep

BW is an important economic trait in sheep that influences growth and development. Currently, most studies have used a single approach to screen genes associated with BW traits in sheep. To address this limitation, we conducted a genome-wide association study (GWAS) covering four different BW periods: birth, weaning, 6 months, and 12 months. Five new candidate genes: MAP3K1, ANKRD55, ABCB1, MEF2C and TRNAW-CCA-87 were screened using a combination of GWAS and quantitative trait loci analysis in sheep. Additionally, five genes were subjected to Gene ontology and Kyoto Encyclopedia of Genes and Genomes pathway enrichment analyses. These genes were primarily enriched in pathways related to growth hormone and energy metabolism. The results demonstrated that the above genes potentially influenced the growth and development of sheep. The five new candidate genes are closely related to the BW trait in sheep, which will be valuable for understanding the genetic mechanisms underlying BW traits and for guiding sheep breeding.

Genetic Structure and Genome-Wide Association Analysis of Growth and Reproductive Traits in Fengjing Pigs

The Fengjing pig is one of the local pig breed resources in China and has many excellent germplasm characteristics. However, research on its genome is lacking. To explore the degree of genetic diversity of the Fengjing pig and to deeply explore its excellent traits, this study took Fengjing pigs as the research object and used the Beadchip Array Infinium iSelect-96|XT KPS_PorcineBreedingChipV2 for genotyping. We analyzed the genetic diversity, relatedness, inbreeding coefficient, and population structure within the Fengjing pig population. Our findings revealed that the proportion of polymorphic markers (PN) was 0.469, and the effective population size was 6.8. The observed and expected heterozygosity were 0.301 and 0.287, respectively. The G-matrix results indicated moderate relatedness within the population, with certain individuals exhibiting closer genetic relationships. The NJ evolutionary tree classified Fengjing boars into five family lines. The average inbreeding coefficient based on ROH was 0.318, indicating a high level of inbreeding. GWAS identified twenty SNPs significantly associated with growth traits (WW, 2W, and 4W) and reproductive traits (TNB and AWB). Notably, WNT8B, RAD21, and HAO1 emerged as candidate genes influencing 2W, 4W, and TNB, respectively. Genes such as WNT8B were verified by querying the PigBiobank database. In conclusion, this study provides a foundational reference for the conservation and utilization of Fengjing pig germplasm resources and offers insights for future molecular breeding efforts in Fengjing pigs.

Elucidation of the genetic determination of body weight and size in Chinese local chicken breeds by large-scale genomic analyses

BACKGROUND

Body weight and size are important economic traits in chickens. While many growth-related quantitative trait loci (QTLs) and candidate genes have been identified, further research is needed to confirm and characterize these findings. In this study, we investigate genetic and genomic markers associated with chicken body weight and size. This study provides new insights into potential markers for genomic selection and breeding strategies to improve meat production in chickens.

METHODS

We performed whole-genome resequencing of and Wenshang Barred (WB) chickens (n = 596) and three additional breeds with varying body sizes (Recessive White (RW), WB, and Luxi Mini (LM) chickens; (n = 50)). We then used selective sweeps of mutations coupled with genome-wide association study (GWAS) to identify genomic markers associated with body weight and size.

RESULTS

We identified over 9.4 million high-quality single nucleotide polymorphisms (SNPs) among three chicken breeds/lines. Among these breeds, 287 protein-coding genes exhibited positive selection in the RW and WB populations, while 241 protein-coding genes showed positive selection in the LM and WB populations. Genomic heritability estimates were calculated for 26 body weight and size traits, including body weight, chest breadth, chest depth, thoracic horn, body oblique length, keel length, pelvic width, shank length, and shank circumference in the WB breed. The estimates ranged from 0.04 to 0.67. Our analysis also identified a total of 2,522 genome-wide significant SNPs, with 2,474 SNPs clustered around two genomic regions. The first region, located on chromosome 4 (7.41-7.64 Mb), was linked to body weight after ten weeks and body size traits. LCORL, LDB2, and PPARGC1A were identified as candidate genes in this region. The other region, located on chromosome 1 (170.46-171.53 Mb), was associated with body weight from four to eighteen weeks and body size traits. This region contained CAB39L and WDFY2 as candidate genes. Notably, LCORL, LDB2, and PPARGC1A showed highly selective signatures among the three breeds of chicken with varying body sizes.

CONCLUSION

Overall this study provides a comprehensive map of genomic variants associated with body weight and size in chickens. We propose two genomic regions, one on chromosome 1 and the other on chromosome 4, that could helpful for developing genome selection breeding strategies to enhance meat yield in chickens.

Genome-wide single nucleotide polymorphism (SNP) data reveal potential candidate genes for litter traits in a Yorkshire pig population

The litter trait is one of the most important economic traits, and increasing litter size is of great economic value in the pig industry. However, the molecular mechanisms underlying pig litter traits remain elusive. To identify molecular markers and candidate genes for pig litter traits, a genome-wide association study (GWAS) and selection signature analysis were conducted in a Yorkshire pig population. A total of 518 producing sows were genotyped with Illumina Porcine SNP 50 BeadChip, and 1969 farrowing records for the total number born (TNB), the number born alive (NBA), piglets born dead (PBD), and litter weight born alive (LWB) were collected. Then, a GWAS was performed for the four litter traits using a repeatability model. Based on the estimated breeding values (EBVs) of TNB, 15 high- and 15 low-prolificacy individuals were selected from the 518 sows to implement selection signature analysis. Subsequently, the selection signatures affecting the litter traits of sows were detected by using two methods including the fixation index (FST) and θ π . Combining the results of the GWAS and selection signature analysis, 20 promising candidate genes (NKAIN2, IGF1R, KISS1R, TYRO3, SPINT1, ADGRF5, APC2, PTBP1, CLCN3, CBR4, HPF1, FAM174A, SCP2, CLIC1, ZFYVE9, SPATA33, KIF5C, EPC2, GABRA2, and GABRA4) were identified. These findings provide novel insights into the genetic basis of pig litter traits and will be helpful for improving the reproductive performances of sows in pig breeding.

A fast non-parametric test of association for multiple traits

The increasing availability of multidimensional phenotypic data in large cohorts of genotyped individuals requires efficient methods to identify genetic effects on multiple traits. Permutational multivariate analysis of variance (PERMANOVA) offers a powerful non-parametric approach. However, it relies on permutations to assess significance, which hinders the analysis of large datasets. Here, we derive the limiting null distribution of the PERMANOVA test statistic, providing a framework for the fast computation of asymptotic p values. Our asymptotic test presents controlled type I error and high power, often outperforming parametric approaches. We illustrate its applicability in the context of QTL mapping and GWAS.

Genome-wide association study of multiyear dynamic growth traits in hybrid Liriodendron identifies robust genetic loci associated with growth trajectories

The genetic factors underlying growth traits differ over time points or stages. However, most current studies of phenotypes at single time points do not capture all loci or explain the genetic differences underlying growth trajectories. Hybrid Liriodendron exhibits obvious heterosis and is widely cultivated, although its complex genetic mechanism underlying growth traits remains unknown. A genome-wide association study (GWAS) is an effective method for elucidating the genetic architecture by identifying genetic loci underlying complex quantitative traits. In the present study, using a GWAS, we identified robust loci associated with growth trajectories in hybrid Liriodendron populations. We selected 233 hybrid progenies derived from 25 crosses for resequencing, and measured their tree height (H) and diameter at breast height (DBH) for 11 consecutive years; 192 972 high-quality single nucleotide polymorphisms (SNPs) were obtained. The dynamics of the multiyear single-trait GWAS showed that year-specific SNPs predominated, and only five robust SNPs for DBH were identified in at least three different years. Multitrait GWAS analysis with model parameters as latent variables also revealed 62 SNPs for H and 52 for DBH associated with the growth trajectory, displaying different biomass accumulation patterns, among which four SNPs exerted pleiotropic effects. All identified SNPs also exhibited temporal variations in effect sizes and inheritance patterns potentially related to different growth and developmental stages. The haplotypes resulting from these significant SNPs might pyramid favorable loci, benefitting the selection of superior genotypes. The present study provides insights into the genetic architecture of dynamic growth traits and lays a basis for future molecular-assisted breeding.

Uncovering the candidate genes related to sheep body weight using multi-trait genome-wide association analysis

In sheep, body weight is an economically important trait. This study sought to map genetic loci related to weaning weight and yearling weight. To this end, a single-trait and multi-trait genome-wide association study (GWAS) was performed using a high-density 600 K single nucleotide polymorphism (SNP) chip. The results showed that 43 and 56 SNPs were significantly associated with weaning weight and yearling weight, respectively. A region associated with both weaning and yearling traits (OARX: 6.74-7.04 Mb) was identified, suggesting that the same genes could play a role in regulating both these traits. This region was found to contain three genes (TBL1X, SHROOM2 and GPR143). The most significant SNP was Affx-281066395, located at 6.94 Mb (p = 1.70 × 10-17), corresponding to the SHROOM2 gene. We also identified 93 novel SNPs elated to sheep weight using multi-trait GWAS analysis. A new genomic region (OAR10: 76.04-77.23 Mb) with 22 significant SNPs were discovered. Combining transcriptomic data from multiple tissues and genomic data in sheep, we found the HINT1, ASB11 and GPR143 genes may involve in sheep body weight. So, multi-omic anlaysis is a valuable strategy identifying candidate genes related to body weight.

Longitudinal genome-wide association studies of milk production traits in Holstein cattle using whole-genome sequence data imputed from medium-density chip data

Longitudinal traits, such as milk production traits in dairy cattle, are featured by having phenotypic values at multiple time points, which change dynamically over time. In this study, we first imputed SNP chip (50-100K) data to whole-genome sequence (WGS) data in a Chinese Holstein population consisting of 6,470 cows. The imputation accuracies were 0.88 to 0.97 on average after quality control. We then performed longitudinal GWAS in this population based on a random regression test-day model using the imputed WGS data. The longitudinal GWAS revealed 16, 39, and 75 quantitative trait locus regions associated with milk yield, fat percentage, and protein percentage, respectively. We estimated the 95% confidence intervals (CI) for these quantitative trait locus regions using the logP drop method and identified 581 genes involved in these CI. Further, we focused on the CI that covered or overlapped with only 1 gene or the CI that contained an extremely significant top SNP. Twenty-eight candidate genes were identified in these CI. Most of them have been reported in the literature to be associated with milk production traits, such as DGAT1, HSF1, MGST1, GHR, ABCG2, ADCK5, and CSN1S1. Among the unreported novel genes, some also showed good potential as candidate genes, such as CCSER1, CUX2, SNTB1, RGS7, OSR2, and STK3, and are worth being further investigated. Our study provided not only new insights into the candidate genes for milk production traits, but also a general framework for longitudinal GWAS based on random regression test-day model using WGS data.

Bayesian analysis of longitudinal traits in the Korea Association Resource (KARE) cohort

Various methodologies for the genetic analysis of longitudinal data have been proposed and applied to data from large-scale genome-wide association studies (GWAS) to identify single nucleotide polymorphisms (SNPs) associated with traits of interest and to detect SNP-time interactions. We recently proposed a grid-based Bayesian mixed model for longitudinal genetic data and showed that our Bayesian method increased the statistical power compared to the corresponding univariate method and well detected SNP-time interactions. In this paper, we further analyze longitudinal obesity-related traits such as body mass index, hip circumference, waist circumference, and waist-hip ratio from Korea Association Resource data to evaluate the proposed Bayesian method. We first conducted GWAS analyses of cross-sectional traits and combined the results of GWAS analyses through a meta-analysis based on a trajectory model and a random-effects model. We then applied our Bayesian method to a subset of SNPs selected by meta-analysis to further discover SNPs associated with traits of interest and SNP-time interactions. The proposed Bayesian method identified several novel SNPs associated with longitudinal obesity-related traits, and almost 25% of the identified SNPs had significant p-values for SNP-time interactions.

A New Algorithm for Multivariate Genome Wide Association Studies Based on Differential Evolution and Extreme Learning Machines

Genome-wide association studies (GWAS) are observational studies of a large set of genetic variants, whose aim is to find those that are linked to a certain trait or illness. Due to the multivariate nature of these kinds of studies, machine learning methodologies have been already applied in them, showing good performance. This work presents a new methodology for GWAS that makes use of extreme learning machines and differential evolution. The proposed methodology was tested with the help of the genetic information (370,750 single-nucleotide polymorphisms) of 2049 individuals, 1076 of whom suffer from colorectal cancer. The possible relationship of 10 different pathways with this illness was tested. The results achieved showed that the proposed methodology is suitable for detecting relevant pathways for the trait under analysis with a lower computational cost than other machine learning methodologies previously proposed.

A genome-wide association study of the longitudinal course of executive functions

Executive functions are metacognitive capabilities that control and coordinate mental processes. In the transdiagnostic PsyCourse Study, comprising patients of the affective-to-psychotic spectrum and controls, we investigated the genetic basis of the time course of two core executive subfunctions: set-shifting (Trail Making Test, part B (TMT-B)) and updating (Verbal Digit Span backwards) in 1338 genotyped individuals. Time course was assessed with four measurement points, each 6 months apart. Compared to the initial assessment, executive performance improved across diagnostic groups. We performed a genome-wide association study to identify single nucleotide polymorphisms (SNPs) associated with performance change over time by testing for SNP-by-time interactions using linear mixed models. We identified nine genome-wide significant SNPs for TMT-B in strong linkage disequilibrium with each other on chromosome 5. These were associated with decreased performance on the continuous TMT-B score across time. Variant rs150547358 had the lowest P value = 7.2 × 10^-10 with effect estimate beta = 1.16 (95% c.i.: 1.11, 1.22). Implementing data of the FOR2107 consortium (1795 individuals), we replicated these findings for the SNP rs150547358 (P value = 0.015), analyzing the difference of the two available measurement points two years apart. In the replication study, rs150547358 exhibited a similar effect estimate beta = 0.85 (95% c.i.: 0.74, 0.97). Our study demonstrates that longitudinally measured phenotypes have the potential to unmask novel associations, adding time as a dimension to the effects of genomics.

Hierarchical mixed-model expedites genome-wide longitudinal association analysis

A hierarchical random regression model (Hi-RRM) was extended into a genome-wide association analysis for longitudinal data, which significantly reduced the dimensionality of repeated measurements. The Hi-RRM first modeled the phenotypic trajectory of each individual using a RRM and then associated phenotypic regressions with genetic markers using a multivariate mixed model (mvLMM). By spectral decomposition of genomic relationship and regression covariance matrices, the mvLMM was transformed into a multiple linear regression, which improved computing efficiency while implementing mvLMM associations in efficient mixed-model association expedited (EMMAX). Compared with the existing RRM-based association analyses, the statistical utility of Hi-RRM was demonstrated by simulation experiments. The method proposed here was also applied to find the quantitative trait nucleotides controlling the growth pattern of egg weights in poultry data.

GASVeM: A New Machine Learning Methodology for Multi-SNP Analysis of GWAS Data Based on Genetic Algorithms and Support Vector Machines

Genome-wide association studies (GWAS) are observational studies of a large set of genetic variants in an individual’s sample in order to find if any of these variants are linked to a particular trait. In the last two decades, GWAS have contributed to several new discoveries in the field of genetics. This research presents a novel methodology to which GWAS can be applied to. It is mainly based on two machine learning methodologies, genetic algorithms and support vector machines. The database employed for the study consisted of information about 370,750 single-nucleotide polymorphisms belonging to 1076 cases of colorectal cancer and 973 controls. Ten pathways with different degrees of relationship with the trait under study were tested. The results obtained showed how the proposed methodology is able to detect relevant pathways for a certain trait: in this case, colorectal cancer.

Rapid epistatic mixed-model association studies by controlling multiple polygenic effects

SUMMARY

We have developed a rapid mixed model algorithm for exhaustive genome-wide epistatic association analysis by controlling multiple polygenic effects. Our model can simultaneously handle additive by additive epistasis, dominance by dominance epistasis and additive by dominance epistasis, and account for intrasubject fluctuations due to individuals with repeated records. Furthermore, we suggest a simple but efficient approximate algorithm, which allows the examination of all pairwise interactions in a remarkably fast manner of linear with population size. Simulation studies are performed to investigate the properties of REMMAX. Application to publicly available yeast and human data has showed that our mixed model-based method has similar performance with simple linear model on computational efficiency. It took less than 40 h for the pairwise analysis of 5000 individuals genotyped with roughly 350 000 SNPs with five threads on Intel Xeon E5 2.6 GHz CPU.

AVAILABILITY AND IMPLEMENTATION

Source codes are freely available at https://github.com/chaoning/GMAT.

SUPPLEMENTARY INFORMATION

Supplementary data are available at Bioinformatics online.

Heterophylly Quantitative Trait Loci Respond to Salt Stress in the Desert Tree Populus euphratica

Heterophylly, or leaf morphological changes along plant shoot axes, is an important indicator of plant eco-adaptation to heterogeneous microenvironments. Despite extensive studies on the genetic control of leaf shape, the genetic architecture of heterophylly remains elusive. To identify genes related to heterophylly and their associations with plant saline tolerance, we conducted a leaf shape mapping experiment using leaves from a natural population of Populus euphratica. We included 106 genotypes grown under salt stress and salt-free (control) conditions using clonal seedling replicates. We developed a shape tracking method to monitor and analyze the leaf shape using principal component (PC) analysis. PC1 explained 42.18% of the shape variation, indicating that shape variation is mainly determined by the leaf length. Using leaf length along shoot axes as a dynamic trait, we implemented a functional mapping-assisted genome-wide association study (GWAS) for heterophylly. We identified 171 and 134 significant quantitative trait loci (QTLs) in control and stressed plants, respectively, which were annotated as candidate genes for stress resistance, auxin, shape, and disease resistance. Functions of the stress resistance genes ABSCISIC ACIS-INSENSITIVE 5-like (ABI5), WRKY72, and MAPK3 were found to be related to many tolerance responses. The detection of AUXIN RESPONSE FACTOR17-LIKE (ARF17) suggests a balance between auxin-regulated leaf growth and stress resistance within the genome, which led to the development of heterophylly via evolution. Differentially expressed genes between control and stressed plants included several factors with similar functions affecting stress-mediated heterophylly, such as the stress-related genes ABC transporter C family member 2 (ABCC2) and ABC transporter F family member (ABCF), and the stomata-regulating and reactive oxygen species (ROS) signaling gene RESPIRATORY BURST OXIDASE HOMOLOG (RBOH). A comparison of the genetic architecture of control and salt-stressed plants revealed a potential link between heterophylly and saline tolerance in P. euphratica, which will provide new avenues for research on saline resistance-related genetic mechanisms.

HiGwas: how to compute longitudinal GWAS data in population designs

SUMMARY

Genome-wide association studies (GWAS), particularly designed with thousands and thousands of single-nucleotide polymorphisms (SNPs) (big p) genotyped on tens of thousands of subjects (small n), are encountered by a major challenge of p ≪ n. Although the integration of longitudinal information can significantly enhance a GWAS's power to comprehend the genetic architecture of complex traits and diseases, an additional challenge is generated by an autocorrelative process. We have developed several statistical models for addressing these two challenges by implementing dimension reduction methods and longitudinal data analysis. To make these models computationally accessible to applied geneticists, we wrote an R package of computer software, HiGwas, designed to analyze longitudinal GWAS datasets. Functions in the package encompass single SNP analyses, significance-level adjustment, preconditioning and model selection for a high-dimensional set of SNPs. HiGwas provides the estimates of genetic parameters and the confidence intervals of these estimates. We demonstrate the features of HiGwas through real data analysis and vignette document in the package.

AVAILABILITY AND IMPLEMENTATION

https://github.com/wzhy2000/higwas.

CONTACT

rwu@phs.psu.edu.

SUPPLEMENTARY INFORMATION

Supplementary data are available at Bioinformatics online.

Genome-wide association study for genotype by lactation stage interaction of milk production traits in dairy cattle

Substantial evidence demonstrates that the genetic background of milk production traits changes during lactation. However, most GWAS for milk production traits assume that genetic effects are constant during lactation and therefore might miss those quantitative trait loci (QTL) whose effects change during lactation. The GWAS for genotype by lactation stage interaction are aimed at explicitly detecting the QTL whose effects change during lactation. The purpose of this study was to perform GWAS for genotype by lactation stage interaction for milk yield, lactose yield, lactose content, fat yield, fat content, protein yield, and somatic cell score to detect QTL with changing effects during lactation. For this study, 19,286 test-day records of 1,800 first-parity Dutch Holstein cows were available and cows were genotyped using a 50K SNP panel. A total of 7 genomic regions with effects that change during lactation were detected in the GWAS for genotype by lactation stage interaction. Two regions on Bos taurus autosome (BTA)14 and BTA19 were also significant based on a GWAS that assumed constant genetic effects during lactation. Five regions on BTA4, BTA10, BTA11, BTA16, and BTA23 were only significant in the GWAS for genotype by lactation stage interaction. The biological mechanisms that cause these changes in genetic effects are still unknown, but negative energy balance and effects of pregnancy may play a role. These findings increase our understanding of the genetic background of lactation and may contribute to the development of better management indicators based on milk composition.