David Rouquié

The paper presents a model validation framework for molecular property prediction models in drug discovery, revealing four failure modes and showing that common evaluation protocols overestimate performance.

Cell Painting in primary human hepatocytes exposed to 1,085 compounds detects bioactivity at lower concentrations than cytotoxicity assays and predicts cell-based, but not cell-free, ToxCast endpoints across feature extractors.

The paper validates a generative AI model using experimental assays to design small molecule drugs targeting specific cancer-related phenotypes, showing 88% morphological impact and 37% transcriptomic pathway effects.

The paper analyzes Cell Painting and transcriptomics data to explore how cell morphology and gene expression change over time and across cell lines in response to compounds.

The paper evaluates read-across methods for predicting rat acute oral toxicity using chemical structure and Cell Painting data, finding chemical structure is more predictive for similar compounds while morphology is better for dissimilar ones.