In microbiome and gene expression studies, sequencing depth is typically arbitrary and does not reflect the biological systems’ size (e.g., total microbial load) being studied. This problem is well known and is the impetus for using normalizations when analyzing these data. Previously, we showed that normalizations make strong, implicit, and often unreasonable biological assumptions that should be avoided. Two viable approaches remain: Scale Invariant and Scale Reliant Inference. In this talk, I will discuss recent advances in both types of modeling, introduce the core concepts behind these approaches, and introduce tools available to researchers looking to perform these analyses.
If it involves cool math and is impactful, I am interested. Lately my research has focused on uncertainty quantification in the setting of partial identifiability with a particular application to the analysis of multivariate sequence count data (e.g., Microbiome and Gene Expression studies). I also have particular interest in multivariate time-series analysis which I have applied to a wide variety of problems in finance, epidemiology, and personalized medicine. In my free time I enjoy exploring the outdoors (backpacking, canoing, rock climbing), creating puzzles, and farming.