Mol Jepa

Mol-JEPA multimodal molecular foundation model (Joint Embedding Predictive Architecture). Input: a list of SMILES. Output: molecular embeddings — a CLS token [batch,512] (global summary), per-modality predicted embeddings [batch,12,512], and per-modality latents [batch,13,512]. Modalities fused: cellular effects, binding affinity, ADMET properties, quantum chemistry. MutGlyph... no — maftools counterpart: none (foundation model).

Use in DarwinOmics

How to use it

Use this model when the submitted analysis depends on embeddings, predictions, scoring, generation, or model-backed comparison. Model output is treated as a reproducible analysis step, not as independent validation of a biological claim.