My research in bioinformatics and computational biology is focused on developing methods that leverage the growing public genomics resources from projects such as TCGA, ICGC and other large-scale genomic projects. I develop methods for integration of genomics data including DNA, RNA and proteomics data (Meng et al., 2014, Fagan et al., 2007, Culhane et al., 2003) and have developed a bi-clustering method called iBBiG, to analyze gene set enrichment analyses results across studies (Gusenleitner et al., 2012). I also led the GeneSigDB database, in which we curated over 5,000 gene signatures of cancer and disease from the published literature (Culhane et al., 2010, Culhane et al., 2012) which can be used for pathway based mining of cancer gene networks. I and my team have created several R packages for integrated exploratory data analysis of multiple genomics datasets including made4 and omicade4. Our R/Bioconductor package MOGSA includes methods for multi-way decomposition of multiple omics data including co-inertia analysis, multiple factor analysis, Generalized canonical correlations analysis and consensus principal component analysis (Meng et al., 2014, Meng et al., doi.org/10.1101/046904). These transform multiple types of genomics data on the same scale, allowing one to extract correlated patterns between data sets, and integrate data for better performance in downstream pathway and geneset analysis.
I lead the Bioinformatics Navigator service, which matches members of DF/HCC and CFAR with bioinformaticians and computational biologists. I am a member of biostatistician/bioinformatics core team for the DF/HCC Renal Cancer SPORE. I also work on women's cancer and am co-PI on a project to study the breast metastatic environment in ER+ breast cancer.