Software

Software Developed and Maintained by the Yang Lab

SR-TWAS can leverage base imputation models for molecular traits that are trained with multiple statistical methods, multiple tissues, or multiple cohorts, to learn an optimal imputation model for the follow-up xWAS analysis.

OTTERS is a powerful xWAS tool leveraging summary-level reference molecular QTL data to conduct gene-based association study with GWAS summary data.

BFGWAS_QUANT conducts Bayesian functional GWAS using multivariate quantitative functional annotations, which can be applied with both individual-level and summary-level GWAS data.

Bayesian Genome-wide TWAS (BGW-TWAS) method is developed for leveraging both cis- and trans- eQTL through summary statistics for TWAS which is based on the BFGWAS framework.

Transcriptome-Integrated Genetic Association Resource (TIGAR) is developed for integrating gene-expression imputation model training, prediction, and TWAS in the same tool. Efficient handling of VCF genotype files and parallele computing have been implemented in TIGAR. TIGAR can efficiently train both Elastic-Net and nonparametric Bayesian DPR models for gene expression imputation.

BFGWAS / BFGWAS_SS

BFGWAS is developed for integrating functional annotation in GWAS by a scalable Bayesian method based on the Bayesian variable selection regression (BVSR) model and Expectation Maximization Monte Carlo Markov Chain (EM-MCMC) algorithm. BFGWAS can now take summary GWAS data.

BFDA is a MATLAB package for Bayesian functional data analysis, especially for smoothing multiple functional samples from the same underlying distribution.