Biophysical modeling with variational autoencoders for bimodal, single-cell RNA sequencing data.

We motivate and present , which combines the variational autoencoder framework of with biophysically motivated, bivariate models for nascent and mature RNA distributions. While previous approaches to integrate bimodal data via the variational autoencoder framework ignore the causal relationship between measurements, models the biophysical processes that give rise to observations. We demonstrate through simulated benchmarking that captures cell type structure in a low-dimensional space and accurately recapitulates parameter values and copy number distributions. On biological data, provides a scalable route for identifying the biophysical mechanisms underlying gene expression. This analytical approach outlines a generalizable strategy for treating multimodal datasets generated by high-throughput, single-cell genomic assays.