MODE: Learning compositional representations of complex systems with Mixtures Of Dynamical Experts
This work addresses a key problem in computational biology for modeling cellular dynamics, offering a novel approach to predict cell fate transitions, though it is incremental in improving upon existing graphical modeling techniques.
The paper tackles the challenge of modeling complex dynamical systems with overlapping behavioral regimes and noisy transitions by proposing MODE, a graphical modeling framework that decomposes dynamics into interpretable components, enabling unsupervised discovery of regimes and accurate long-term forecasting, with demonstrated success on synthetic and real datasets including human single-cell RNA sequencing data.
Dynamical systems in the life sciences are often composed of complex mixtures of overlapping behavioral regimes. Cellular subpopulations may shift from cycling to equilibrium dynamics or branch towards different developmental fates. The transitions between these regimes can appear noisy and irregular, posing a serious challenge to traditional, flow-based modeling techniques which assume locally smooth dynamics. To address this challenge, we propose MODE (Mixture Of Dynamical Experts), a graphical modeling framework whose neural gating mechanism decomposes complex dynamics into sparse, interpretable components, enabling both the unsupervised discovery of behavioral regimes and accurate long-term forecasting across regime transitions. Crucially, because agents in our framework can jump to different governing laws, MODE is especially tailored to the aforementioned noisy transitions. We evaluate our method on a battery of synthetic and real datasets from computational biology. First, we systematically benchmark MODE on an unsupervised classification task using synthetic dynamical snapshot data, including in noisy, few-sample settings. Next, we show how MODE succeeds on challenging forecasting tasks which simulate key cycling and branching processes in cell biology. Finally, we deploy our method on human, single-cell RNA sequencing data and show that it can not only distinguish proliferation from differentiation dynamics but also predict when cells will commit to their ultimate fate, a key outstanding challenge in computational biology.