Accelerating New Drug Discovery

What this post added

This post details the integration of Milvus with RDKit for high-performance chemical structure similarity search in drug discovery. It explains how RDKit generates Morgan fingerprints from chemical structures (SMILES format) and how these fingerprints are imported into Milvus for substructure, similarity, and exact searches. The post highlights the use of Tanimoto distance (implemented as Jaccard distance in Milvus) for chemical fingerprint computation and demonstrates a system demo with over 90 million chemical fingerprints.