Recursion Pharmaceuticals

A self-supervised masked autoencoder for RNA-seq count data, pretrained on 1.4M public samples to learn transferable transcriptomic representations without per-dataset re-training.

Open model that jointly predicts biomolecular structure and small-molecule binding affinity, approaching FEP+ accuracy in seconds on a single GPU.

Channel-agnostic Vision Transformer trained on 3M+ Cell Painting images via masked autoencoder, producing 384-dimensional morphological embeddings for zero-shot phenotypic analysis.