The Cell Maps for Artificial Intelligence (CM4AI) project seeks to map the spatiotemporal architecture of human cells and use these maps to enable interpretable genotype-to-phenotype learning. In genomics and precision medicine, machine learning models often function as “black boxes,” predicting phenotypes from genotypes without revealing the biological mechanisms behind those predictions.

CM4AI aims to overcome this limitation through a coordinated, multimodal effort that integrates proteomic mass spectrometry, cellular imaging, and genetic perturbation using CRISPR/Cas9. These complementary approaches will generate a large-scale library of cellular maps capturing protein function, organization, and gene regulatory effects across diverse demographic and disease contexts. By anchoring AI models in biological mechanism, CM4AI supports the development of more transparent and trustworthy applications in genomic medicine.