*********************************
There is now a CONTENT FREEZE for Mercury while we switch to a new platform. It began on Friday, March 10 at 6pm and will end on Wednesday, March 15 at noon. No new content can be created during this time, but all material in the system as of the beginning of the freeze will be migrated to the new platform, including users and groups. Functionally the new site is identical to the old one. webteam@gatech.edu
*********************************
Ann Chen, Ph.D.
Department of Biostatistics and Bioinformatics
Moffitt Cancer Center
Abstract
Single-cell technologies allow characterization of genomics, transcriptomes, and epigenomes for individual cells under different conditions and provide unprecedented resolution for researchers. We will first introduce an interactive toolbox SinCHet, which we develop to analyze single cell data for studying heterogeneity using Shannon Profile of at different resolutions. A novel D statistic using area under the Profile of Shannon Differences is devised to detect heterogeneity differences between conditions. Recently, we generalize this tool by implementing de-batching and subpopulation-comparison modules in SinCHet-MS for analyzing single cell mass spectrometry (SCMS) metabolomics data. These suites of tools provide insights into emerging or disappearing subpopulations between conditions, and enable the prioritization of biomarkers for follow-up experiments based on heterogeneity or marker differences between and/or within sub-populations. Two datasets will be discussed during the first part of the talk. The first dataset is a single cell mRNA (scRNA) dataset from two melanoma cell lines and mouse models. The analyses show that melanomas consisted of multiple transcriptional states that they have different drug sensitivities and growth dynamics under drug. The heterogeneity analyses further showed that tumor size in the melanoma mouse model is negatively associated with transcriptional diversity. The second dataset is a SCMS dataset from two colon cell lines. Although unbiased profiling is powerful, we showed that initial experimental design with careful de-batching first is still essential to gain biological insights from single cell data.
As data dimensionality increase quickly, the number cells and genes for scRNA-seq quickly rise to tens of thousands easily (e.g., 10X genomics), analyses become the rate-limiting step. The combination of parameters for nonlinear models is large to investigate. We employ a JavaScript-based solution, Single Cell Visual Analytics (SCVA), to create an enriched and fast online environment to allow the experts interactively investigate various aspects of their single cell data, including exploring combinations of different parameters for t-SNE projections in real time, cell type recognition, and tumor-environment change in response to treatments in patient samples. Analyses of scRNA-seq analyses in patient samples collected on baseline and day 8 on-treatment tumor biopsies show increase in immune cell influx (CD4+ and CD8+ T-cells) and a decrease in number of cancer cells on treatment.
About the Speaker
Dr. Chen’s research has been focused on developing statistical methods and computational tools to incorporate multiple omics sources, select biologically relevant markers, and predict clinical outcomes in a unified framework. Her work on Bayesian methodological development of data integration for regulatory network inference and pathway and gene selection for cancer survival prediction facilitates the identification of deregulated pathways with therapeutic relevance in subsets of human cancer. Dr. Chen’s work on nonparametric method improvement for the detection of nonlinear correlation has enabled the identification of key genes for the development of pathological conditions, which might have been missed by traditional methods to detect merely linear relationships. Dr. Chen’s recent work is focused on developing methods to use next gen- sequencing and other omics data to identify novel targetable pathways for melanoma patients, especially for those who did not have commonly known driver mutations.
