Cimplrx

CIMPLRX is a drug discovery company based in Anyang-si, Gyeonggi-do, Korea, that discovers and develops new drug candidates using its own explainable artificial intelligence (XAI) platform, CEEK-CURE. The platform is split into two components: CEEK (Create ExplainablE Knowledge), the user-facing interface used to analyze and visualize results and capture the discovery process, and CURE (Collect and Uncover RElationships), the server-side XAI engine that handles requests coming from CEEK.

The platform covers several early discovery workflows. Target-based hit identification takes a UniProt ID as input, collects known compounds associated with that target, applies supervised and unsupervised learning, and uses the resulting models to design new compounds and screen commercial and public databases automatically. Compound-based hit identification starts from seed compounds, defines biological spaces using six different chemical spaces, collects known compound-target relationships, and builds directional and predictive models for automated compound design. Lead optimization uses prebuilt models to make predictions and filter compounds, with colored and sized circles overlaid on chemical structures to indicate the direction and strength of each substructure's contribution. Mechanism-of-action identification uses active compounds to build multiple predictive models, reporting the top three predicted targets per compound, their frequencies, and comparisons between input compounds and compounds in the training set.

CEEK-CURE is organized around explaining models, results and processes: users can select and combine models and view predictions with colored structures, prediction values and standard deviations; results can be examined through parallel coordinates, chord diagrams, dendrograms, heatmaps, interactive scatter plots and clustering and diversity tools for structure-activity relationship analysis; and the drug development process is displayed as a graph navigable via protein nodes or keyword search, with links to UniProt for active sites, compounds and interaction data and PyMOL for viewing protein active sites. The company also describes an approach aimed at the earliest phase of drug discovery, finding new compounds while simultaneously identifying new uses for existing drugs.

Business model

CIMPLRX both discovers and develops its own new drug candidates and makes its AI platform available for drug discovery work, listing universities, biotech and pharmaceutical companies, and investment institutions among its partner categories.