Large-Scale Survey of Cell-Differentiation Programs in a Generative Model Reveals Regeneration as an Epiphenomenon of Development

TL;DR

This study uses a large-scale generative model to analyze cell-differentiation programs, revealing that cell lineage graphs are more likely directed acyclic graphs with convergence, and that regeneration is a common outcome of development.

Contribution

The paper introduces a novel generative modeling approach to explore the statistical properties of developmental schemes and cell differentiation trajectories.

Findings

Cell-type lineage graphs are more likely DAGs than trees.

Most generated organisms can regenerate using pluripotent cells.

Developmental schemes exhibit converging lineages rather than diverging trees.

Abstract

Development combines three basic processes asymmetric --- cell division, signaling and gene regulation --- in a multitude of ways to create an overwhelming diversity of multicellular life-forms. Here, we attempt to chart this diversity using a generative model. We sample millions of biologically feasible developmental schemes, allowing us to comment on the statistical properties of cell-differentiation trajectories they produce. Our results indicate that, in contrast to common views, cell-type lineage graphs are unlikely to be tree-like. Instead, they are more likely to be directed acyclic graphs, with multiple lineages converging on the same terminal cell-type. Additionally, in line with the hypothesis that whole body regeneration is an epiphenomenon of development, a majority of the `organism' generated by our model can regenerate using pluripotent cells. The generative framework is modular and flexible, and can be adapted to test additional hypotheses about general features of development.