Azbukina, N. | Search bioRxiv

bioRxiv2025

CellFlow enables generative single-cell phenotype modeling with flow matching

High-content phenotypic screens provide a powerful strategy for studying biological systems, but the scale of possible perturbations and cell states makes exhaustive experiments unfeasible. Computational models that are trained on existing data and extrapolate to correctly predict outcomes in unseen contexts have the potential to accelerate biological discovery. Here, we present CellFlow, a flexible framework based on flow matching that can model single cell phenotypes induced by complex perturbations. We apply CellFlow to various phenotypic screens, accurately predicting expression responses to a wide range of perturbations, including cytokine stimulation, drug treatments and gene knockouts. CellFlow successfully modeled developmental perturbations at the whole-embryo scale and guided cell fate and organoid engineering by predicting heterogeneous cell populations arising from combinatorial morphogen treatments and by performing a virtual organoid protocol screen. Taken together, CellFlow has the potential to accelerate discovery from phenotypic screens by learning from existing data and generating phenotypes induced by unseen conditions.

bioinformatics↗

bioRxiv2025

Multi-omic human neural organoid cell atlas of the posterior brain

Patterning of the neural tube establishes midbrain and hindbrain structures that coordinate motor movement, process sensory input, and integrate cognitive functions. Cellular impairment within these structures underlie diverse neurological disorders, and in vitro organoid models promise inroads to understand development, model disease, and assess therapeutics. Here, we use paired single-cell transcriptome and accessible chromatin sequencing to map cell composition and regulatory mechanisms in organoid models of midbrain and hindbrain. We find that existing midbrain organoid protocols generate ventral and dorsal cell types, and cover regions including floor plate, dorsal and ventral midbrain, as well as adjacent hindbrain regions, such as cerebellum. Gene regulatory network (GRN) inference and transcription factor perturbation resolve mechanisms underlying neuronal differentiation. A single-cell multiplexed patterning screen identifies morphogen concentration and combinations that expand existing organoid models, including conditions that generate medulla glycinergic neurons and cerebellum glutamatergic subtypes. Differential abundance of cell states across screen conditions enables differentiation trajectory reconstruction from region-specific progenitors towards diverse neuron types of mid- and hindbrain, which reveals morphogen-regulon regulatory relationships underlying neuronal fate specification. Altogether, we present a single-cell multi-omic atlas and morphogen screen of human neural organoid models of the posterior brain, advancing our understanding of the co-developmental dynamics of regions within the developing human brain.

developmental biology↗