scGeneScope: A Treatment-Matched Single Cell Imaging and Transcriptomics Dataset and Benchmark for Treatment Response Modeling

Joel Dapello, Marcel Nassar, Ridvan Eksi, Ban Wang, Jules Gagnon-Marchand, Kenneth T Gao, Akram Baharlouei, Kyra Thrush-Evensen, Nina Riehs, Amy F Peterson, Aniket Tolpadi, Abhejit Rajagopal, Henry E Miller, Ashley Mae Conard, David Alvarez-Melis, Rory Stark, Simone Bianco, Morgan Levine, Ava P Amini, Alex Xijie Lu et al. (4 additional authors not shown)

Keywords: scRNA-seq, Cell Painting microscopy, multimodal, multisample, chemical perturbation, data, benchmark, foundation models

TL;DR: We create a perturbationally-paired single cell imaging and transcriptomics dataset and benchmark and evaluate unimodal, multimodal, and multisample imaging and scRNA-seq models for treatment response modeling; HF data access token in zip.

Abstract: Understanding cellular responses to chemical interventions is critical to the discovery of effective therapeutics. Because individual biological techniques often measure only one axis of cellular response at a time, high-quality multimodal datasets are needed to unlock a holistic understanding of how cells respond to treatments and to advance computational methods that integrate modalities. However, many techniques destroy cells and thus preclude paired measurements, and attempts to match disparate unimodal datasets are often confounded by data being generated in incompatible experimental settings. Here we introduce scGeneScope, a multimodal single‑cell RNA sequencing (scRNA-seq) and Cell Painting microscopy image dataset conditionally paired by chemical treatment, designed to facilitate the development and benchmarking of unimodal, multimodal, and multiple profile machine learning methods for cellular profiling. 28 chemicals, each acting on distinct biological pathways or mechanisms of action (MoAs), were applied to U2-OS cells in two experimental data generation rounds, creating paired sets of replicates that were then profiled independently by scRNA‑seq or Cell Painting. Using scGeneScope, we derive a replicate- and experiment-split treatment identification benchmark simulating MoA discovery under realistic laboratory variability conditions and evaluate unimodal, multimodal, and multiprofile models ranging in complexity from linear approaches to recent foundation models. Multiprofile integration improved performance in both the unimodal and multimodal settings, with gains more consistent in the former. Evaluation of unimodal models for MoA identification demonstrated that recent scRNA-seq foundation models deployed zero-shot were consistently outperformed by classic fit-to-data methods, underscoring the need for careful, realistic benchmarking in machine learning for biology. We release the scGeneScope dataset and benchmarking code to support further research.

Dataset URL: https://huggingface.co/datasets/altoslabs/scGeneScope

Primary Area: AL/ML Datasets & Benchmarks for life sciences (e.g. climate, health, life sciences, physics, social sciences)

Submission Number: 877