The latent simplicity of biological teams

Abstract

The functions of biological systems emerge from both the individual contributions of their components and the interactions among them. Even in modestly sized biological systems, the number of such interactions is enormous, scaling exponentially with the number of components. Exhaustively characterizing all interactions is generally infeasible, hampering our ability to model complex biological systems. Using microbial communities as our primary case study, we find that despite their large potential complexity and high dimensionality, ecological function is often remarkably simple: it is governed by as few as one or two collective modes and is predictable from a small set of effective interaction parameters. We show that these collective modes can be explained mechanistically and emerge naturally from shared constraints that render interactions non-independent. We highlight how this insight enables the construction of predictive, interpretable, and quantitative models of microbial consortia. Our theoretical framework is general and extends to a broad range of systems, from protein fitness landscapes to human team performance.

Grupo Interdisciplinar de Sistemas Complejos