Abstract
Individual cells in a population generally have different replicative capability, presumably due to the phenotypic variability of the cells. Identifying the latent states that rule the replicative capability and characterizing how the states are inherited over generations are crucial for understanding how the self-replication of the cells is modulated and controlled for achieving higher fitness and resistance to different kinds of perturbations. Even with technological development to monitor the proliferation of single cells over tens of generations and to trace the lineages of cells, estimating the state of the cells is still hampered by the lack of statistical methods that can appropriately account for the lineage specific problems. In this work, we develop a statistical method to infer the growth-related latent states of cells over a cellular lineage tree concurrently with the switching dynamics of the states and the statistical law how the state determines the division time. An application of our method to a lineage data of E.coli has identified a three dimensional effective state in the cells, one component of which seems to capture slow fluctuation of cellular state over generations.
Footnotes
SN, YS, and TJK designed and performed research; SN and TJK analyzed data; SN, YS, and TJK wrote the paper.
The authors declare no conflict of interest.