The Frost Lab develops bioinformatics methods for analyzing high-dimensional genomic data with an emphasis on single cell and spatial transcriptomics. We have created a diverse set of single cell methods including gene set testing techniques, methods for inferring cell type and cell-surface receptor abundance, and approaches for characterizing cell-cell interactions.
More theoretical research in statistics and applied mathematics provides an important foundation for our bioinformatics projects. Research interests in this space include sparse matrix decomposition, randomized numerical linear algebra, hypothesis aggregation/weighting, penalized estimation, network analysis, hypercomplex numbers, and random matrix theory.
Although most of our research is methodological, we actively collaborate with multiple experimental research groups on projects involving the generation and analysis of single cell data with study of the tumor microenvironment an important focus. Other translational areas of interest include tissue-specific gene activity, gene-environment and gene-gene interactions, cancer prognosis prediction, and the human microbiome.
Image details: False color visualization of a quaternion representation of mouse kidney spatial transcriptomics data. Details can be found in a bioRxiv preprint at https://doi.org/10.1101/2022.07.21.501020.