PubMed Health⌕ Search

SEARCH · PubMed Health

Results for “genotype by environment interactions”

Explore indexed PubMed citations for clinical trials, systematic reviews and public health research. Read source abstracts and follow each citation to its original PubMed record.

Quote a phrase for an exact phrase match. Source license links do not imply unrestricted reuse.

At least 55 records · Page 3Linked to original sources
PubMed1975

The inheritance of scholastic abilities in a sample of twins. II. Genetical analysis of examinations results.

1. Examination results for 98 pairs of same-sex and 39 pairs of opposite-sex twins were available for analysis as well as measurements of IQ and height for 47 of the same sex pairs. 2. The choice of subjects taken for examination is not influenced by genetic differences within DZ pairs but seems to be mainly determined by between family cultural differences. 3. Neither the sex difference in opposite-sex pairs nor the difference between first and second born twins in same-sex pairs contributes appreciably to within pair environmental variance. However, same-sex DZ pairs are more often found in different school classes than MZ pairs. 4. No evidence for interaction between genotypes and within-family environments was found for any character. 5. The simple environmental model fails to fit the data for English, French, Geography, Mathematics and IQ whereas the simple genetical model fits. For the remaining subjects it is not possible to exclude either hypothesis. The heritablities of the examination performances are of the same magnitude as that for IQ. 6. Sampling inadequacies are revealed in the data for height but there is evidence for a substantial genetics component of phenotype variation. 7. There is evidence for heterogeneity of gene action in the subjects English and Mathematics supporting the view that there are genetically determined specific abilities acting independently of general intelligence.

Adolescent↗
PubMed2026

Genetic and environmental interactions outweigh mitonuclear coevolution for complex traits in Drosophila.

The interdependent relationship between mitochondrial and nuclear genomes is a powerful model for understanding how epistasis shapes the architecture and evolution of complex traits. Once considered a neutral marker, mitochondrial DNA variation is now recognized as critical to phenotypic evolution because of its epistatic interactions and history of coevolution with the nuclear genome. A central challenge in evolutionary genetics is to quantify the relative importance of stabilizing and directional selection shaping complex trait distributions within and among species. Both can act on interacting and/or co-evolving genes contributing to quantitative traits, but resolving their relative roles is complicated by the complex architecture of most traits. Here, we use a panel of 90 Drosophila mitonuclear genotypes to quantify the relative contributions of mitochondrial, nuclear, and environmental variation and their interactions to four metabolically demanding complex traits. We sample both within-species and between-species mitochondrial variation and observe stronger interaction effects attributable to within-species variation, consistent with stabilizing selection maintaining mitonuclear function. Additionally, culturing the flies on a mitochondrial Complex I inhibitor, rotenone, reveals significant genotype x environment (G×E and G×G×E) interaction effects, providing insight into how genetic variation can be maintained across changing environments. Our results have broader implications in medicine, where mitochondrial DNA donors with longer purifying selection histories may be safer for mitochondrial replacement therapies.

Journal Article↗
PubMed2025

Order among chaos: High throughput MYCroplanters can distinguish interacting drivers of host infection in a highly stochastic system.

The likelihood that a host will be susceptible to infection is influenced by the interaction of diverse biotic and abiotic factors. As a result, substantial experimental replication and scalability are required to identify the contributions of and interactions between the host, the environment, and biotic factors such as the microbiome. For example, pathogen infection success is known to vary by host genotype, bacterial strain identity and dose, and pathogen dose. Elucidating the interactions between these factors in vivo has been challenging because testing combinations of these variables quickly becomes experimentally intractable. Here, we describe a novel high throughput plant growth system (MYCroplanters) to test how multiple host, non-pathogenic bacteria, and pathogen variables predict host health. Using an Arabidopsis-Pseudomonas host-microbe model, we found that host genotype and bacterial strain order of arrival predict host susceptibility to infection, but pathogen and non-pathogenic bacterial dose can overwhelm these effects. Host susceptibility to infection is therefore driven by complex interactions between multiple factors that can both mask and compensate for each other. However, regardless of host or inoculation conditions, the ratio of pathogen to non-pathogen emerged as a consistent correlate of disease. Our results demonstrate that high-throughput tools like MYCroplanters can isolate interacting drivers of host susceptibility to disease. Increasing the scale at which we can screen drivers of disease, such as microbiome community structure, will facilitate both disease predictions and treatments for medicine and agricultural applications.

Arabidopsis↗
PubMed2026

Nonuniform Association of Genetic Risk Scores for Intraocular Pressure.

IMPORTANCE: Elevated intraocular pressure (IOP) is a risk factor for primary open-angle glaucoma, and genetic risk scores hold promise as a tool for screening for ocular hypertension. However, genetic risk scores for IOP have a nonuniform association across the range of IOP, which reduces their accuracy. OBJECTIVE: To test the hypothesis that nonuniform behavior of genetic risk scores for IOP is associated with a specific type of genetic interaction. DESIGN, SETTING, AND PARTICIPANTS: Cross-sectional, post hoc genetic association studies were performed using linear and quantile regression in a sample of UK Biobank participants. Data were analyzed from January to September 2025. EXPOSURES: Ninety-eight genetic variants associated with IOP. MAIN OUTCOMES AND MEASURES: Tests were carried out for 98 genetic variants associated with IOP (P&#x2009;<&#x2009;5.0&#x2009;&#xd7;10-8) to examine (1) dominant or recessive genetic effects, (2) genotype&#x2009;&#xd7;&#x2009;genotype interactions, (3) genotype&#x2009;&#xd7;&#x2009;age interactions, and (4) genotype&#x2009;&#xd7;&#x2009;sex interactions. RESULTS: A total of 98&#x202f;235 participants (mean [SD] age, 58.1 [7.9] years; 52&#x202f;168 female [53.1%]) were included in this analysis. More variants exhibited genotype&#x2009;&#xd7;&#x2009;age interactions than expected by chance (14 of the 98 variants associated with IOP had at least nominal evidence of an interaction with age; P&#x2009;=&#x2009;3.76&#x2009;&#xd7;10-4). For 12 of these 14 variants, age increased rather than decreased the magnitude of the IOP vs genotype association. However, integrating age interactions into the genetic risk score construction process did not yield improved accuracy (incremental noninteraction model, R2&#x2009;=&#x2009;4.05; 95% CI, 3.82-4.31 and interaction model, R2&#x2009;=&#x2009;4.04; 95% CI, 3.80-4.27). There was little support for other types of genetic interaction. CONCLUSIONS AND RELEVANCE: In the current work, findings show minimal evidence that nonadditive allelic effects, genotype&#x2009;&#xd7;&#x2009;genotype interactions, and genotype&#x2009;&#xd7;&#x2009;sex interactions contributed to the nonuniform association of genetic variants with IOP across quantiles of IOP. Although a genetic risk score for IOP was more accurate in older vs younger individuals, efforts to account for genotype&#x2009;&#xd7;&#x2009;age interactions in genetic risk score construction did not improve accuracy. These findings suggest other factors, such as gene-environment interactions, contribute to the nonuniform relationship of genetic variants with IOP.

Humans↗
PubMed2026

A cross-sectional study of oxidative stress pathway genotypes and their interactions with environmental pollutant levels identifies associations with gene expression and lung function.

BACKGROUND: Asthma is a heterogeneous disease influenced by genetic and environmental factors. Fine particulate matter (PM2.5) exacerbates asthma, likely through oxidative stress pathways, but whether genetic variation modifies this effect remains unclear. METHODS: We analysed data on 948 adults with asthma from the Severe Asthma Research Program (SARP), linking ZIP-code-level PM2.5 exposure with whole-genome sequencing data. We tested 4337 single nucleotide polymorphisms (SNPs) in 120 oxidative stress pathway genes for gene-environment (GxE) interactions with PM2.5 on lung function (forced expiratory volume in 1 s [FEV1] % predicted) using weighted linear regression. Gene expression data from bronchial epithelial cells (n = 170) were used to assess cis-expression quantitative trait loci (eQTLs). FINDINGS: Higher PM2.5 exposure was associated with lower FEV1% predicted (&#x3b2; per &#x3bc;g/m3 = -0.7, p = 0.01). We identified 20 SNPs across seven genes (OXSR1, PXDN, TPO, LRRK2, APP, MSRA, MSRB2) with significant GxE interactions after multiple-testing correction. Five SNPs were also eQTLs, linking PM2.5-modified gene expression to lung function. Minor alleles in OXSR1 and PXDN were associated with reduced gene expression and worsened FEV1% under high PM2.5 exposure. Conversely, TPO variants were associated with higher baseline expression and lower lung function, but under increasing PM2.5 exposure, minor allele carriers showed suppressed TPO expression and improved FEV1%. INTERPRETATION: This study identified 20 SNPs in oxidative stress pathway genes that modify the effect of PM2.5 on lung function in asthma. These findings highlight the importance of integrating environmental context in genetic studies and suggest potential therapeutic targets for pollution-sensitive asthma phenotypes. FUNDING: Supported by NIH grants.

Cross-Sectional Studies↗
PubMed1979

Comparison of antigen-specific I-region-associated cell interaction factors.

Two basic types of factors reacting with anti-I region (anti-Ia) antisera are compared, those derived from macrophage-like antigen presenting cells and others derived from T-lymphocytes, of either the suppressor or helper type. Despite the common property of reacting with anti-Ia antisera, the two sets of factors differ by many criteria. Macrophages, upon culture with antigen, release complexes of Ia antigen and a fragment of the original immunogen. This material is only produced by responder macrophages and thus appears to be a soluble Ir gene product. The genetic restriction of the T-macrophage interaction was investigated in chimeras, and it was found that the host environment as well as the donor genotype was of importance in determining restrictions, which were thus not really directed to "self." There was no evidence for intrinsic T-cell Ir genes, as nonresponder stem cells developed into responder T-cells in a (responder X nonresponder) F1 environment. However, these cells only responded in the presence of responder macrophages. Specific T-cell factors are different in nature. These all react with anti-Ia antisera, but the nature or function of the T-cell Ia is unknown. The basic structure involves a VARIAble region" responsible for antigen binding which, as it reacts with anti-idiotype antisera and anti-variable region framework antisera is an immunoglobulin variable region. There is also a "constant region," defined by its biological properties as well as by specific rabbit antisera. This two-region nature of specific factors is reminiscent of immunoglobulin structure and it is a reasonable hypothesis that the constant region is linked to the Ig cluster of genes.

Animals↗
PubMed2026

Upper airway microbiome interacts with GSDMB and ORMDL3 asthma risk SNPs to influence early-life wheeze risk.

BACKGROUND: Single-nucleotide polymorphisms (SNPs) in the chromosome 17q12-q21 region and, independently, early-life nasal microbiota dominated by Moraxella, Streptococcus, or Haemophilus (MSH) increase risk of chronic wheeze and asthma development. OBJECTIVE: We sought to determine whether 17q12-q21 risk SNPs and nasal microbiota interact to modulate childhood wheeze risk. METHODS: Nasal wash samples from 12-month-old infants in 2 birth cohorts, COAST (Childhood Origins of Asthma; n = 180) and URECA (Urban Environment and Childhood Asthma; n = 139), underwent 16S ribosomal RNA variable region 4 sequencing. Nasal microbiota dominated by MSH or Corynebacterium, Dolosigranulum, Staphylococcus, or Bacillus (CDSB) were assessed. Paired blood was genotyped for 9 17q12-q21 risk SNPs. Logistic regression tested interactions between 17q12-q21 SNPs and MSH or CDSB on wheeze risk in the first 3 years of life. A549 lung epithelial cells, CRISPR-edited to encode the rs7216389 risk genotype (rs7216389TT) were compared to the heterozygous (rs7216389CT) line using bulk RNA sequencing. RESULTS: SNPs, particularly those in the ORMDL3 (rs8076131; odds ratio [OR]: 1.72; 95% CI: 1.09-2.71; Pint = .031) and GSDMB (rs2305480; OR: 1.72; 95% CI: 1.09-2.71; Pint = 0.042; and rs7216389; OR: 1.73; 95% CI: 1.09-2.70; Pint = .047) genes, interact with MSH microbiota to increase early-life wheeze risk (false discovery rate Pint = .016 for all), while interactions with CDSB reduce risk. A549 airway epithelial cells homozygous for rs7216389TT exhibited decreased expression of genes involved in antimicrobial responses and neutrophil recruitment and evidence increased microbial adherence compared with the heterozygous cell line. CONCLUSION: Airway microbiota interact with SNPs at the 17q12-q21 locus in genes involved in sphingolipid metabolism and intracellular antimicrobial responses, to modulate wheeze risk.

Humans↗
PubMed2025

Deciphering the Impact of Temperature on Pleiotropic Consequences of RNA Polymerase Mutations.

Despite occurring in an essential molecule, mutations in RNA polymerase readily emerge and elicit complex pleiotropic effects across different levels of biological organization, which are all modulated by environment. We investigated the impact of temperature on the effects of six mutations on sequence, structure, transcriptome, and organismal traits. We found temperature altered the transcriptomic response and key organismal traits such as growth rate and biofilm formation in a genotype-specific manner. Critically, mechanistic insights into the possible drivers of mutational effects emerged only when examining the relationships between different levels of organization: location of mutations in the tertiary structure and distance to key interacting molecules partly explained the observed transcriptomic differences, which in turn drove the impact of mutations on organismal traits. While falling short of capturing the full complexity of the system, our findings underscore the benefits of integrating insights across multiple biological levels to understand the relationship between environment and mutational effects in molecules with extensive pleiotropic effects.

Mutation↗
PubMed2025

Foliar disease resistance phenomics of fungal pathogens: image-based approaches for mapping quantitative resistance in cereal germplasm.

Host plant resistance is the most effective and environmentally sustainable means of reducing yield losses caused by fungal foliar pathogens of cereal species. Cereal genebank collections hold diverse pools of potentially underutilized disease resistance alleles, and cereal genomic resources are well advanced due to large-scale sequencing and genotyping efforts. Genome-Wide Association Studies (GWAS) have emerged as the predominant association genetics technique to initially discover novel disease resistance loci or alleles in these diverse collections. Traditional disease resistance phenotyping methods are reliant on visual estimation of disease symptom severity and have successfully supported genetic mapping studies either via GWAS or QTL mapping in biparental populations facilitating both marker development and gene cloning efforts. Due to foliar pathogens having a high capacity to evolve, there is a need to pyramid disease resistance genes with diverse mechanisms for durable control. Resistance expressed as a quantitative trait, known as quantitative resistance (QR), is hypothesized to be more durable, unlike major R-gene resistance that is race-specific and can be vulnerable to breaking down without gene stewardship. However, assessing QR visually is challenging, particularly when complicated by complex genotype&#x2009;&#xd7;&#x2009;environment (G&#x2009;&#xd7;&#x2009;E) effects in the field. High-throughput image-based phenotyping provides accurate and unbiased data that can support foliar disease resistance screening efforts of genebank collections using GWAS. In this review, we discuss image-based disease phenotyping based on macroscopic (visible symptoms) and microscopic features during the host-pathogen interaction. Quantitative image analysis approaches using conventional and artificial intelligence (AI) algorithms are also discussed.

Disease Resistance↗
PubMed2026

Gene-environment interaction between perinatal oxytocin exposure and Pten mutation shapes epigenetic reprogramming of oxytocin signaling and behavior in mice.

Synthetic oxytocin (Pitocin) is the most commonly used pharmacologic agent for induction and augmentation of labor. Beyond its uterotonic effects, oxytocin plays a critical role in neurodevelopment and social behavior. Dysregulated oxytocin signaling has been implicated in autism spectrum disorder (ASD), raising concern that perinatal exposure to exogenous oxytocin may have lasting neurodevelopmental consequences. This study aimed to determine whether offspring harboring a genetic predisposition for ASD are differentially impacted by perinatal oxytocin exposures, with a focus on long-term oxytocin signaling and autism-like behavior. Pregnant mice carrying offspring with heterozygous mutations in phosphatase and tensin homolog deleted on chromosome ten (Pten), a well-established monogenic risk factor for ASD, received continuous oxytocin versus phosphate-buffered saline (PBS) control via micro-osmotic pumps during late gestation. Wild-type (WT) offspring exposed to each treatment served as a secondary control. Adult offspring were assessed for oxytocin receptor (Oxtr) methylation in the frontal cortex and hippocampus, oxytocin expression in the hypothalamus, serum oxytocin levels, and were subject to a battery of social and anxiety-related behavior tests. Perinatal oxytocin exposure produced genotype-dependent effects in offspring. Epigenetic analyses revealed bidirectional remodeling of Oxtr methylation in the frontal cortex and hippocampus, with increased exon 1 methylation in WT mice and decreased methylation in Pten-mutant mice, resulting in significant genotype-treatment interactions. Hypothalamic oxytocin expression increased following treatment regardless of genotype, though baseline levels were higher in Pten-mutant mice. Neither oxytocin treatment nor genotype impacted long-term serum oxytocin levels. Behavioral outcomes were modest but context-specific: repetitive behaviors and cognition performance were unchanged, but oxytocin-treated Pten-mutant mice exhibited increased anxiety-like behavior alongside improved social memory. In contrast, oxytocin-treated WT mice showed reduced social novelty preference. Exploratory analyses suggested potential sex-dependent trends. Our findings support a model in which genetic susceptibility shapes the epigenetic encoding of early-life hormonal signals, thereby recalibrating oxytocin system function and downstream behavioral outcomes. Together, these data highlight the context-dependent effects of perinatal oxytocin exposure and argue against uniformly beneficial or detrimental effects, emphasizing the importance of gene-environment interactions in neurodevelopmental trajectories.

Animals↗
PubMed2026

Combining ability and gene action for grain yield and biofortification traits in pearl millet [Pennisetum glaucum (L.) R. Br.]: implications for breeding high-yielding biofortified hybrids in arid regions.

Hybrid RIB-9184 &#xd7; RIB-15131 combines high yield (18.84 g plant&#x207b;&#xb9;) with iron (46.16 mg kg&#x207b;&#xb9;), zinc (38.86 mg kg&#x207b;&#xb9;), and protein (11.91%); Fe-Zn correlation (rg = 0.82) permits simultaneous biofortification. Pearl millet [Pennisetum glaucum (L.) R. Br., syn. Cenchrus americanus (L.) Morrone] is a climate-resilient cereal with inherently high micronutrient levels, making it a priority crop for biofortification. Understanding gene action for yield and nutritional traits is essential for designing effective breeding strategies. Ten diverse inbred lines were crossed in a half-diallel design (Griffing's Method 2, Model 1), and the 55 entries (45 F1 hybrids + 10 parents) were evaluated across two sowing-date environments in a randomised complete block design with three replications at Jaipur, Rajasthan, India. Biofortification traits (Fe, Zn, protein) showed predominantly additive gene action (Baker's ratio 0.71-0.91) with high heritability (0.90-0.94). G&#xd7;E interaction was significant for Fe and Zn but genotypic variance was substantially larger, maintaining high heritability; protein showed no G&#xd7;E interaction. Grain yield was governed largely by non-additive effects (Baker's ratio 0.54) with significant G&#xd7;E interaction, favouring hybrid breeding. Among parents, RIB-9205 had the highest GCA for Fe (6.65, P&#x2009;<&#x2009;0.001), RIB-9184 for Zn (3.85, P&#x2009;<&#x2009;0.001) and protein (0.78, P&#x2009;<&#x2009;0.001), and RIB-9185 was a balanced combiner for yield (1.39, P&#x2009;<&#x2009;0.001) and micronutrients. The hybrid RIB-9184 &#xd7; RIB-15131 ranked first across all five weighting schemes of the multi-trait performance index (1.31), combining grain yield of 18.84&#xa0;g plant&#x207b;1 with Fe of 46.16&#xa0;mg&#xa0;kg&#x207b;1, Zn of 38.86&#xa0;mg&#xa0;kg&#x207b;1, and protein of 11.91%. The strong Fe-Zn correlation (rg = 0.82, P&#x2009;<&#x2009;0.01) permits simultaneous micronutrient improvement. An integrated approach combining hybrid development for yield with population improvement for micronutrient density is recommended for biofortified pearl millet cultivars in arid regions.

Pennisetum↗
PubMed2026

Cross-Kingdom Siderophores: Biosynthesis, Ecology, and Biotechnological Applications.

Microbial siderophores are high-affinity iron-binding compounds which are produced by bacteria, fungi, and actinomycetes to obtain iron and survive and interact with different species in an iron-deficient environment. While the conventional research on siderophore systems deals mainly with the study within the same taxa, modern researchers have increased their inclination toward cross-kingdom integration of siderophore behavior and their impact on host-associated environments. This can be largely attributed to differences in biosynthetic gene clusters, receptor systems, and regulatory networks, which produce distinct genotype-to-phenotype results determining microbial cooperation and competition. Current advancements in genomic research, together with omics studies like transcriptomics, proteomics, and metabolomics, have created newer insights into how siderophores function. However, the present literature evidences multiple major gaps in multi-omics data because the link between genomes and metabolomes remains weak due to inconsistent regulatory data sets and failure in identifying producer-consumer relationships in polymicrobial systems. Additionally, major constraints like molecular instability, delivery system limitations, host toxicity, limitations in upscaling, and regulatory issues delimit the use of siderophores in medical treatment, agricultural practices, and environmental biotechnology. This review aims to bridge the existing knowledge about siderophore biochemistry, biosynthesis, ecological functions, and genetic regulation across kingdoms while integrating multi-omics outlook with translational considerations. Thus, by connecting molecular mechanisms with evolutionary cross-talk, this study aims to provide a system-level framework in the world of siderophore-mediated iron uptake and therefore shapes future directions in emerging fields of microbial engineering, precision therapies, and sustainable biotechnology.

Fur regulation↗
PubMed2026

Life in sediments fosters 'sexual' speciation in the Shewanella baltica complex.

Understanding how intra- and interspecific differentiation arises in natural microbial populations is central to explaining the processes that drive bacterial evolution. Motivated by the co-occurrence of multiple putative genospecies closely related to Shewanella baltica in Baltic Sea sediments, we investigated the genomic structure of this species complex across fine spatial scales. We analyzed 112 genome sequences from strains collected across several sediment cores and depths (0-6 cm) at Vax&#xf6;n (Stockholm archipelago, Sweden) as well as earlier isolates from this site and allopatric strains from surrounding locations obtained from both sediments and the water column. Using a reverse-ecology population genomics approach, we found unprecedented genomic diversification among sediment-associated strains, which form a species complex resolving into three cohesive evolutionary groups (G1, G2, and G3) with distinct signatures of metabolic specialization including sulfite respiration. While G1 consists predominantly of a single species (S. baltica) with high gene turnover, G2 and G3 comprise an array of divergent putative genospecies and previously reported species consistently recovered from sediments. Patterns of homologous recombination indicate that diversification of the lineages within G2 and G3 is primarily recombination-driven ('sexual') and is associated with specialization in sulfite reduction and utilization of certain carbon sources. The extent of diversity uncovered here far exceeds that reported for S. baltica from other environments, suggesting that a sediment-associated lifestyle promotes the emergence of novel genotypes. These findings expand the known limits of sympatric speciation in prokaryotes beyond subspecific ecotypes, demonstrating that bacterial species can diverge and persist as distinct lineages in the absence of spatial segregation and at microgeographic scales. Furthermore, our results suggest that collective interactions and ecological differentiation can structure sediment-associated bacterial populations strongly enough to drive divergence at the species level.

Journal Article↗
PubMed2025

Sex as a modifier of genetic risk for type 1 diabetes.

Sex differences influence the pathogenesis of type 1 diabetes (T1D), yet most genetic studies have treated sex as a control covariate rather than a dynamic effect modifier. Sex influences immune cell behaviour, including CD4+ and CD8+ T cell activation, regulatory T cell stability, B cell autoantibody production, dendritic cell priming and monocyte/macrophage inflammation. Underlying mechanisms include hormone-responsive enhancers, X-escape gene dosage and sex-biassed chromatin states, intersecting with T1D-associated variants to produce sex-specific immune phenotypes. These insights help explain regional variation in sex ratios of T1D incidence, such as male predominance in high-risk populations and female excess in low-risk populations. Biological sex shapes T1D risk across multiple layers, including polygenic load; environmental exposures such as vitamin D deficiency and enteroviral infection; and sex-specific hormonal, chromosomal and epigenetic influences. An integrative G&#x2009;&#xd7;&#x2009;E&#x2009;&#xd7;&#x2009;S (genetic&#x2009;&#xd7;&#x2009;environmental&#x2009;&#xd7;&#x2009;sex-specific) liability-threshold framework is thus supported. Clinical and translational implications include developing sex-specific polygenic risk scores, biomarker panels and interventional strategies targeting pathways such as hormone signalling, vitamin D metabolism and the microbiome. Future multi-omic, longitudinal studies are warranted to test genotype-sex interactions, integrate sex as a core effect modifier and enable precision prevention and treatment of T1D in both males and females.

Humans↗
PubMed2026

Genomic selection in timothy (Phleum pratense L.): a comprehensive evaluation of prediction models, multi-trait strategies, and forward validation across Norwegian environments.

This study presents a comprehensive evaluation of genomic selection (GS) in timothy (Phleum pratense L.), comparing nine prediction models across yield and quality traits at two Norwegian locations. Forward validation with independent full-sib (FS2) families revealed a substantial generalization gap, highlighting the need for realistic accuracy assessment in polyploid forage breeding. Timothy (Phleum pratense L.) is the most important forage grass in Northern Europe, yet genomic selection has not been systematically evaluated in this hexaploid species. We assessed 889 FS2-families originating from biparental crosses among 49 cultivars/populations. The FS2-families were genotyped with 30,698 SNP markers derived from genotyping-by-sequencing (GBS) and field tested for three harvest years at a highland and a lowland continental location in Southern Norway. Nine genomic prediction models were compared for six yield traits (dry matter yield per cut and total) and six quality traits (protein, digestibility, and fiber fractions) across three cuts/year. Within-training cross-validation accuracies were moderate to high (mean r = 0.62), with Random Forest and SVR consistently outperforming GBLUP. However, forward validation using 213 independent FS2-families revealed dramatically lower accuracies (mean r = 0.16), with only 16 of 30 trait-dataset combinations reaching statistical significance (p < 0.05). Genomic heritabilities (GREML), estimated across environments, ranged from near zero for the quality traits to 0.55 for the yield traits. Multi-trait models improved accuracy by 3-5% over single-trait approaches, while FS2 families-by-environment interaction models with Random Forest achieved the highest within-training accuracy (mean r = 0.71). Marker density analysis showed accuracy plateauing at approximately 15000 SNPs. Genetic correlations among the yield component traits were estimated by multi-trait REML; correlations among the quality traits could not be estimated reliably because their genomic heritabilities were low. A multi-trait selection index identified top-performing FS2-families for further crossing recommendations. These results provide a benchmark for GS implementation in hexaploid timothy and emphasize that cross-validation substantially overestimates prediction accuracy for truly independent material.

Norway↗