Revisiting genome-wide association studies from statistical modelling to machine learning.

Over the last decade, genome-wide association studies (GWAS) have discovered thousands of genetic variants underlying complex human diseases and agriculturally important traits. These findings have been utilized to dissect the biological basis of diseases, to develop new drugs, to advance precision medicine and to boost breeding. However, the potential of GWAS is still underexploited due to methodological limitations. Many challenges have emerged, including detecting epistasis and single-nucleotide polymorphisms (SNPs) with small effects and distinguishing causal variants from other SNPs associated through linkage disequilibrium. These issues have motivated advancements in GWAS analyses in two contrasting cultures-statistical modelling and machine learning. In this review, we systematically present the basic concepts and the benefits and limitations in both methods. We further discuss recent efforts to mitigate their weaknesses. Additionally, we summarize the state-of-the-art tools for detecting the missed signals, ultrarare mutations and gene-gene interactions and for prioritizing SNPs. Our work can offer both theoretical and practical guidelines for performing GWAS analyses and for developing further new robust methods to fully exploit the potential of GWAS.

Quantitative trait locus (QTL) analysis and fine-mapping for Fusarium oxysporum disease resistance in Raphanus sativus using GRAS-Di technology.

Fusarium wilt is a significant disease in radish, but the genetic mechanisms controlling yellows resistance (YR) are not well understood. This study aimed to identify YR-QTLs and to fine-map one of them using F2:3 populations developed from resistant and susceptible radish parents. In this study, two high-density genetic maps each containing shared co-dominant markers and either female or male dominant markers that spanned 988.6 and 1127.5 cM with average marker densities of 1.40 and 1.53 cM, respectively, were generated using Genotyping by Random Amplicon Sequencing-Direct (GRAS-Di) technology. We identified two YR-QTLs on chromosome R2 and R7, and designated the latter as ForRs1 as the major QTL. Fine mapping narrowed down the ForRs1 locus to a 195 kb region. Among the 16 predicted genes in the delimited region, 4 genes including two receptor-like protein and -kinase genes (RLP/RLK) were identified as prime candidates for ForRs1 based on the nucleotide sequence comparisons between the parents and their predicted functions. This study is the first to use a GRAS-Di for genetic map construction of cruciferous crops and fine map the YR-QTL on the R7 chromosome of radish. These findings will provide groundbreaking insights into radish YR breeding and understanding the genetics of YR mechanism.

Pathogenic and antimicrobial resistance genes in Streptococcus oralis strains revealed by comparative genome analysis.

Streptococcus oralis is an early colonizer bacterium in dental plaques and is considered a potential pathogen of infective endocarditis (IE) disease. In this study, we built a complete genome map of Streptococcus oralis strain SOT, Streptococcus oralis strain SOD and Streptococcus infantis strain SO and performed comparative genomic analysis among these three strains. The results showed that there are five genomic islands (GIs) in strain SOT and one CRISPR in strain SOD. Each genome harbors various pathogenic genes related to diseases and drug resistance, while the antibiotic resistance genes in strains SOT and SOD were quite similar but different from those in strain SO. In addition, we identified 17 main virulence factors and capsule-related genes in three strains. These results suggest the pathogenic potential of Streptococcus strains, which lay a foundation for the prevention and treatment of a Streptococcus oralis infection.

Genetic mapping using 1.4M SNP array refined loci for fatty acid composition traits in Chinese Erhualian and Bamaxiang pigs.

Chinese indigenous pigs display marked genetic and phenotypic differences compared with western commercial pigs. In this study, we tested the association between 660K SNPs and longissimus muscle fatty acid composition traits in Chinese Erhualian (n = 331) and Bamaxiang (n = 315) pigs based on a customized 1.4 million SNP array. We identified a total of 64 significant associations for 20 fatty acid composition traits at the p-value threshold of 1 × 10-6 among which 42 associations in low linkage disequilibrium (r2  < .2) with previously reported loci were considered novel. We substantially improved the strength and precision of the associations at four previously detected loci near FADS2, ELOVL7, ELOVL6 and FASN genes, facilitating follow-up candidate gene studies. Moreover, we also identified loci near ABCD2, ACSBG1, ELOVL5, HPGDS, DAGT2, ACAD10 and ACSL1 genes with function relevant to metabolism of fatty acids. In this study, valuable genetic variants and candidate genes associated with fatty acid composition traits were identified in Erhualian and Bamaxiang pigs. Some identified loci could be used to improve pork nutrition in pig breeding practice. Using the SNP array with higher marker density and less ascertainment bias improved QTL detection power and precision in Chinese indigenous pigs.

Marker development using SLAF-seq and whole-genome shotgun strategy to fine-map the semi-dwarf gene ari-e in barley.

BACKGROUND: Barley semi-dwarf genes have been extensively explored and widely used in barley breeding programs. The semi-dwarf gene ari-e from Golden Promise is an important gene associated with some agronomic traits and salt tolerance. While ari-e has been mapped on barley chromosome 5H using traditional markers and next-generation sequencing technologies, it has not yet been finely located on this chromosome. RESULTS: We integrated two methods to develop molecular markers for fine-mapping the semi-dwarf gene ari-e: (1) specific-length amplified fragment sequencing (SLAF-seq) with bulked segregant analysis (BSA) to develop SNP markers, and (2) the whole-genome shotgun sequence to develop InDels. Both SNP and InDel markers were developed in the target region and used for fine-mapping the ari-e gene. Linkage analysis showed that ari-e co-segregated with marker InDel-17 and was delimited by two markers (InDel-16 and DGSNP21) spanning 6.8 cM in the doubled haploid (DH) Dash × VB9104 population. The genetic position of ari-e was further confirmed in the Hindmarsh × W1 DH population which was located between InDel-7 and InDel-17. As a result, the overlapping region of the two mapping populations flanked by InDel-16 and InDel-17 was defined as the candidate region spanning 0.58 Mb on the POPSEQ physical map. CONCLUSIONS: The current study demonstrated the SLAF-seq for SNP discovery and whole-genome shotgun sequencing for InDel development as an efficient approach to map complex genomic region for isolation of functional gene. The ari-e gene was fine mapped from 10 Mb to 0.58 Mb interval.

Fine mapping of leaf delayed virescence gene dv4 in Triticum aestivum.

Wheat (Triticum aestivum L.) is one of the most important crops worldwide, and its yield affects national food security. Wheat leaves are key photosynthetic organs where carbohydrates are synthesized for grain yield. Leaf colour mutants are ideal germplasm resources for molecular genetic studies of wheat chloroplast development, chlorophyll synthesis and photosynthesis. We obtained a wheat mutant delayed virescence 4 (dv4) from cultivar Guomai 301. The leaves of mutant dv4 were pale yellow at the seedling stage, golden yellow at the turning green stage, and they started to turn green at the jointing stage. Genetic analysis demonstrated that the yellow-leaf phenotype was controlled by a single recessive gene named as dv4. Gene dv4 was fine mapped in a 1.46 Mb region on chromosome 7DS by SSR and dCAPS marker assays. Three putative candidate genes were identified in this region. Because no leaf colour genes have been reported on wheat chromosome arm 7DS previously, dv4 is a novel leaf colour gene. The result facilitates map-based cloning of dv4 and provides information for the construction of a high-photosynthetic efficiency ideotype for improving wheat yield.

A DNA Regulatory Element Haplotype at Zinc Finger Genes Is Associated with Host Resilience to Small Ruminant Lentivirus in Two Sheep Populations.

Small ruminant lentivirus (SRLV) causes Maedi-Visna or Ovine Progressive Pneumonia in sheep and creates insidious livestock production losses. This retrovirus is closely related to human immunodeficiency virus and currently has no vaccines or cure. Genetic marker assisted selection for sheep disease resiliency presents an attractive management solution. Previously, we identified a region containing a cluster of zinc finger genes that had association with ovine SRLV proviral concentration. Trait-association analysis validated a small insertion/deletion variant near ZNF389 (rs397514112) in multiple sheep breeds. In the current study, 543 sheep from two distinct populations were genotyped at 34 additional variants for fine mapping of the regulatory elements within this locus. Variants were selected based on ChIP-seq annotation data from sheep alveolar macrophages that defined active cis-regulatory elements predicted to influence zinc finger gene expression. We present a haplotype block of variants within regulatory elements that have improved associations and larger effect sizes (up to 4.7-fold genotypic difference in proviral concentration) than the previously validated ZNF389 deletion marker. Hypotheses for the underlying causal mutation or mutations are presented based on changes to in silico transcription factor binding sites. These variants offer alternative markers for selective breeding and are targets for future functional mutation assays.

Fine-Mapping and Analysis of Cgl1, a Gene Conferring Glossy Trait in Cabbage (Brassica oleracea L. var. capitata).

Cuticular waxes covering the outer plant surface impart a whitish appearance. Wax-less cabbage mutant shows glossy in leaf surface and plays important roles in riching cabbage germplasm resources and breeding brilliant green cabbage. This is the first report describing the characterization and fine-mapping of a wax biosynthesis gene using a novel glossy Brassica oleracea mutant. In the present paper, we identified a glossy cabbage mutant (line10Q-961) with a brilliant green phenotype. Genetic analyses indicated that the glossy trait was controlled by a single recessive gene. Preliminary mapping results using an F2 population containing 189 recessive individuals revealed that the Cgl1 gene was located at the end of chromosome C08. Several new markers closely linked to the target gene were designed according to the cabbage reference genome sequence. Another population of 1,172 recessive F2 individuals was used to fine-map the Cgl1 gene to a 188.7-kb interval between the C08SSR61 simple sequence repeat marker and the end of chromosome C08. There were 33 genes located in this region. According to gene annotation and homology analyses, the Bol018504 gene, which is a homolog of CER1 in Arabidopsis thaliana, was the most likely candidate for the Cgl1 gene. Its coding and promoter regions were sequenced, which indicated that the RNA splice site was altered because of a 2,722-bp insertion in the first intron of Bol018504 in the glossy mutant. Based on the FGENESH 2.6 prediction and sequence alignments, the PLN02869 domain, which controls fatty aldehyde decarbonylase activity, was absent from the Bol018504 gene of the 10Q-961 glossy mutant. We inferred that the inserted sequence in Bol018504 may result in the glossy cabbage mutant. This study represents the first step toward the characterization of cuticular wax biosynthesis in B. oleracea, and may contribute to the breeding of new cabbage varieties exhibiting a brilliant green phenotype.