Fate specification and variability in cell lineage trees

Caleb Lau⁰, Terry Speed¹, Sarah Russell², Damien Hicks⁰
(0) Swinburne University of Technology
(1) Walter & Eliza Hall Institue of Medical Research
(2) Peter MacCallum Cancer Centre

Find me on Wed Nov 25th, 1:30-2:50pm AEDT in Remo, table 32

Abstract
Cell lineage trees provide a blueprint for understanding how cell fate is speci- fied. Sulston et al. (1983) famously showed how the lineage tree for C. elegans pinpoints when each cell type is specified during embryonic development. This success was possible largely because the invariant cell lineage tree for C. elegans allows trees from different founder cells to be inspected visually. For more complex organisms, however, cell specification is significantly more stochastic, making it impossible to use such simple visualisation techniques to interpret the lineage tree.

To address this problem, Hicks et al (PLOS Comp Bio 2019) recently developed a new technique, called the lineage variability map, which provides the statistical framework for understanding lineage trees at the population, not just individual, level. This method showed how cell specification should be asso ated with the parts of the tree that are strong sources of variation. The original application of the method to studying the highly variable T-cell (CD8+) lineage tree was, however, limited by the large quantities of incomplete data. Here we show how to solve this problem using a Bayesian analysis that samples from the distribution of missing data as well as from the posterior distribution of the tree parameters. This enables us to extend our analysis of T cells into the regime of significant cell differentiation. .