The Ma Lab at Fred Hutchinson Cancer Center specializes in statistical and computational methods for genomic data, in particular microbiome data. We employ a variety of statistical learning methods, ranging from dimensionality reduction, graphical models, and high-dimensional inference, to address the analytical challenges faced with interpreting omics data. The long-term goals of our research are to enhance biomarker discoveries through powerful and robust statistical inference, and to translate these findings to advance clinical research.
Keywords: microbiome, network analysis, high-dimensional inference, data integration
We have a new preprint on mediation analysis with compositional exposures. The method summarizes the gut microbiome by a latent microbial balance, estimating community-level direct and indirect effects while identifying the taxa that drive them. The R package BalExMed implements the method, and the slides from my JSM 2026 talk give a quick overview.
We have a preprint on Bayesian covariance regression for differential microbial networks. This work was led by former RA student Zichun Xu.
Our paper on inference for microbe-metabolite association networks is out in Biometrics! For a quick overview of the method, also check out my talk.
Our paper on constructing canine comorbidity networks using data from the Dog Aging Project was featured in EurekAlert!! This work was led by a former undergraduate intern Antoinette Fang.
The Section on Statistics in Genomics and Genetics (SSGG) of the American Statistical Association is pleased to announce the 2025 Distinguished Student Paper Award Competition.
For eligibility criteria and application guidelines, please go to https://lnkd.in/ga6G2Ys6
All materials must be received by 11:59 PM (Pacific Time) November 15, 2024.