Masked Autoencoders for Microscopy are Scalable Learners of Cellular Biology
This work addresses the problem of scalable image analysis for cellular biology researchers, offering incremental improvements with potential applications in drug discovery.
The paper tackled the challenge of featurizing microscopy images for large-scale biological experiments by scaling weakly supervised classifiers and self-supervised masked autoencoders (MAEs), showing that ViT-based MAEs achieve up to 11.5% relative improvement in recalling known biological relationships and generalize effectively across different experimental conditions.
Featurizing microscopy images for use in biological research remains a significant challenge, especially for large-scale experiments spanning millions of images. This work explores the scaling properties of weakly supervised classifiers and self-supervised masked autoencoders (MAEs) when training with increasingly larger model backbones and microscopy datasets. Our results show that ViT-based MAEs outperform weakly supervised classifiers on a variety of tasks, achieving as much as a 11.5% relative improvement when recalling known biological relationships curated from public databases. Additionally, we develop a new channel-agnostic MAE architecture (CA-MAE) that allows for inputting images of different numbers and orders of channels at inference time. We demonstrate that CA-MAEs effectively generalize by inferring and evaluating on a microscopy image dataset (JUMP-CP) generated under different experimental conditions with a different channel structure than our pretraining data (RPI-93M). Our findings motivate continued research into scaling self-supervised learning on microscopy data in order to create powerful foundation models of cellular biology that have the potential to catalyze advancements in drug discovery and beyond.