Morphologically Intelligent Perturbation Prediction with FORM

2025-10-24

AI Summary

Current computational models are largely restricted to two-dimensional cell morphology modeling, failing to capture realistic three-dimensional structural dynamics under perturbations—thereby limiting the predictive accuracy and functional interpretability of virtual cells. To address this, we propose the first 3D cell morphology prediction framework supporting both unconditional generation and conditional simulation. Our method employs a multi-channel VQGAN-based high-fidelity morphological encoder coupled with a diffusion model to explicitly model perturbation-induced morphological trajectories, trained on over 65,000 3D multi-fluorescence cellular volume images. The framework accurately predicts 3D morphological evolution under chemical or genetic interventions, infers signaling pathway activity, disentangles combinatorial perturbation effects, and generalizes to unseen perturbation types. We further release MorphoEval, a standardized benchmarking suite for quantitative evaluation, establishing a new paradigm and reproducible evaluation standard for 3D virtual cell modeling.

Technology Category

• Computer Vision: Diffusion Models for Vision
• Machine Learning: Large Multimodal Models (LMMs)
• Cognitive Modeling & Cognitive Systems: Computational Creativity

Application Category

• User Modeling, Personalization and Recommendation: User model development and evaluation
• Graph Algorithms and Modeling for the Web: Foundation models and LLMs for Web-related graphs
• Web Mining and Content Analysis: Models for Web evolution

Abstract

Understanding how cells respond to external stimuli is a central challenge in biomedical research and drug development. Current computational frameworks for modelling cellular responses remain restricted to two-dimensional representations, limiting their capacity to capture the complexity of cell morphology under perturbation. This dimensional constraint poses a critical bottleneck for the development of accurate virtual cell models. Here, we present FORM, a machine learning framework for predicting perturbation-induced changes in three-dimensional cellular structure. FORM consists of two components: a morphology encoder, trained end-to-end via a novel multi-channel VQGAN to learn compact 3D representations of cell shape, and a diffusion-based perturbation trajectory module that captures how morphology evolves across perturbation conditions. Trained on a large-scale dataset of over 65,000 multi-fluorescence 3D cell volumes spanning diverse chemical and genetic perturbations, FORM supports both unconditional morphology synthesis and conditional simulation of perturbed cell states. Beyond generation, FORM can predict downstream signalling activity, simulate combinatorial perturbation effects, and model morphodynamic transitions between states of unseen perturbations. To evaluate performance, we introduce MorphoEval, a benchmarking suite that quantifies perturbation-induced morphological changes in structural, statistical, and biological dimensions. Together, FORM and MorphoEval work toward the realisation of the 3D virtual cell by linking morphology, perturbation, and function through high-resolution predictive simulation.

Problem

• Predicts 3D cellular structural changes under perturbation
• Overcomes limitations of 2D models in capturing morphology complexity
• Links cellular morphology with perturbation effects and function

Innovation

• 3D machine learning framework for cellular morphology prediction
• Multi-channel VQGAN encoder learns compact 3D shape representations
• Diffusion-based module models morphological evolution across perturbations