Activity Number:
|
188
- Bayesian Application to Biological and Health Sciences
|
Type:
|
Contributed
|
Date/Time:
|
Tuesday, August 4, 2020 : 10:00 AM to 2:00 PM
|
Sponsor:
|
Section on Bayesian Statistical Science
|
Abstract #313645
|
|
Title:
|
A Bayesian Latent Feature Model for Phenotypic Heterogeneity in Single-Cell Mass Cytometry Data
|
Author(s):
|
Thomas Madsen* and Franziska Michor
|
Companies:
|
Carleton College and Dana-Farber Cancer Institute
|
Keywords:
|
Bayesian nonparametrics;
Biostatistics;
Cancer Biology
|
Abstract:
|
Phenotypic heterogeneity is a key determinant of cancer progression and the emergence of therapeutic resistance, but meaningful quantification of this heterogeneity remains a challenge. We present a latent feature model which quantifies heterogeneity in single-cell mass cytometry data, permitting comparison of heterogeneity across experimental conditions. We apply a Bayesian matrix factorization approach to learn the underlying biological pathways that generate the data. Simulation studies confirm that this model is able to correctly rank samples in terms of heterogeneity in these underlying features. The model is also able to distinguish a pooled sample of mass cytometry data from thirteen triple-negative breast cancer cell line samples, demonstrating its capacity to identify and rank samples in terms of phenotypic heterogeneity. Our work suggests that latent feature models can provide biologically relevant quantification of heterogeneity, permitting the investigation of therapies which target phenotypic heterogeneity itself.
|
Authors who are presenting talks have a * after their name.