Discovering innovative drugs depends on the ability to search vast chem. diversity efficiently. Carmot's discovery technol., Chemotype Evolution, provides rapid access to chem. space relevant to a target of interest. Chemotype Evolution can swiftly and iteratively generate and screen libraries of small mols. or small mol.-peptide hybrids to produce drug leads. The process starts by designing an anchor mol. or "bait". The bait can be derived from known inhibitors, substrates, co-factors, peptides, hits from a fragment screen, or pharmaceutically acceptable "warheads". In addition to its target-interacting components, the bait contains a reactive functionality such that it can be linked individually with members of Carmot's 16,000+ fragment collection, which has been custom-built for the technol. over several years focusing on diversity, lead-likeness, low mol. weight, and lipophilicity. The linked mols. (each consisting of a bait and a fragment) constitute a custom library biased towards the target. This customized library is screened to detect binding or modulation of target activity in biochem. or cell-based assays. Chemotype Evolution does not screen pools of compounds; all compounds are synthesized and tested individually. Hits identified from a first iteration screen can be fed into a medicinal chem. program or converted into new baits and used in a second iteration screen to identify more potent mols. Alternatively, fragment hits can be repurposed as baits to identify fragments that replace the initial anchor mol. This process can be repeated as many times as desired. Chemotype Evolution is highly efficient, with tens of thousands of mols. screened per iteration, and does not require protein structural information. This presentation will demonstrate how Chemotype Evolution, medicinal chem., and structure-based drug design were combined to discover novel irreversible small mol. inhibitors of the oncogenic G12C mutant form of KRAS that show potent biochem. and cell-based activity.