Methodological innovations for genetic mapping of human diseases
A substantial part of our recent research is developing novel statistics and bioinformatics methods to optimally integrate genetic data and multi-omic resources to identify loci or genes responsible for human disorders.
We have a long-standing interest in developing statistic methods for integrating the emerging genomics, epigenomics and transcriptomics resources into conventional statistical tests to detect sequence variants or genes contributing to human genetic diseases. In the past few years, we have invented a series of statistical and bioinformatics methods in this field. These include a rapid and powerful gene-based statistical association test for genome-wide association studies (GWAS) [Li MX et al. Am J Hum Genet.2011;88(3):283-93], a hybrid statistical test for protein-protein interaction-bases association analysis in GWAS [Li MX et al. Am J Hum Genet. 2012;91(3):478-88], a comprehensive framework to prioritize sequence variants in exome sequencing studies of human diseases [Li MX et al. Nucleic Acids Res. 2012;40(7):e53, 2017;45(9):e75 ], an ultra-fast and accurate method to impute statistical significance at untyped sequence variants [Eur J Hum Genet. 2016;24(5):761-6], a powerful method for estimation of the cancer-driver genes [Jiang L. et al. Nucleic Acids Res. 2019;47(16):e96], an accurate method for estimating driver tissues by selective expression of genes associated with complex diseases or traits[Genome Biol. 2019;20(1):233], etc.
The above methods have been implemented into software tools for integrative genetic mapping. For example, KGGSeq is a widely-used tool with compressive downstream analysis function for detecting mutations and genes of human diseases from high throughput sequencing data. KGG is a tool for combining multiple GWAS association signals for knowledge-based association analysis. KGGSee is a tool for leveraging gene expression profiles to mine causal genes and cell-types from GWAS association signals.