Gene-based GWAS analysis for consecutive studies of GEFOS

Abstract

Purpose Single-nucleotide polymorphism (SNP)-based genome-wide association studies (GWASs) have already identified about 100 loci associated with bone mineral density (BMD), but these loci only explain a small proportion of heritability to osteoporosis risk. In the present study, we performed a gene-based analysis of the largest GWASs in the bone field to identify additional BMD-associated genes. Methods BMD-associated genes were identified by combining the summary statistic P values of SNPs across individual genes in the two consecutive meta-analyses of GWASs from the Genetic Factors for Osteoporosis (GEFOS) studies. The potential functionality of these genes to bone was partially assessed by differential gene expression analysis. Additionally, the consistency of the identification of potential BMD-associated variants were evaluated by estimating the correlation of the P values of the same SNPs/genes between the two consecutive GEFOS studies with largely overlapping samples. Results Compared to the SNP-based analysis, the gene-based strategy identified additional BMD-associated genes with genome-wide significance and increased their mutual replication between the two GEFOS datasets. Among these BMD-associated genes, 3 novel genes (UBTF, AAAS, and C11orf58) were partially validated at the gene expression level. The correlation analysis presented a moderately high between-study consistency of potential BMD-associated variants. Conclusions Gene-based analysis as a supplementary strategy to SNP-based GWAS, when applied here, is shown that it helped identify some novel BMD-associated genes. In addition to its empirically increased statistical power, gene-based analysis also provides a higher testing stability for identification of BMD genes.

A road map for understanding molecular and genetic determinants of osteoporosis

“…Another useful approach is that used in gene-based GWAS, which regard a gene as a basic unit for association analyses, thereby improving the statistical power (BOX 2) by overcoming the potential allelic heterogeneity (BOX 1) and reducing the number of independent statistical tests from millions (for SNPs) to tens of thousands (for genes) 44 . A 2018 study found three novel BMD-associated genes (UBTF, AAAS and C11orf58) through a gene-based GWAS strategy 44 . The roles of the proteins encoded by these three genes in bone homeostasis are still unknown.…”

Machine learning, transcriptome, and genotyping chip analyses provide insights into SNP markers identifying flower color in Platycodon grandiflorus

“…A transcriptome-based SNP approach can be a cost/time-saving method for identifying large-scale markers 37 . We found 9 transcriptome-wide SNPs in the coding regions associated with flower color (Table 2 ).…”

Identification of transcriptome-wide, nut weight-associated SNPs in Castanea crenata

“…A transcriptome-based SNP approach can efficiently address these problems. Additionally, transcriptome-based GWAS represents a cost-effective and time-saving method for developing large-scale markers 16 , especially compared with the general GWAS or other conventional marker identification assays. Recently, transcriptome-based GWAS has been successfully applied to human diseases 16,17 .…”

“…In particular, genome-based breeding reduces the time required for various breeding schemes 14,15 . Furthermore, a transcriptome-based GWAS, which is a type of GWAS, uses transcriptome sequencing data to identify variants within the coding regions of the genome 16 . Transcriptome-based GWAS has several advantages compared with the traditional GWAS.…”

“…Third, transcriptome-based analysis also practically reduces the burden of multiple testing for traditional GWASs. Furthermore, the gene is the functional unit in the genome with high consistency across populations, which is the major target used by most of the subsequent bioinformatics analyses 16 . Machine learning (ML) is the study of algorithms and statistical models that facilitate software applications to increase the accuracy of outcome predictions without explicit instructions.…”