BRAF-Mutant Melanoma Case Study

From a library of 2 million molecules, our platform predicted 1806 compounds with potential activity in BRAF-mutant melanoma. Among the top-ranked compounds, 10 compounds were novel and showed the strongest predicted melanoma-selective responses.

We tested five of these candidates at 50 nM using 48-hour live-cell imaging.
Four produced clear, selective killing of melanoma cells, with minimal effects in control lines – matching the model’s predicted phenotypic differences.

This workflow reduced the typical discovery process to:

• 5 experiments instead of thousands

• 4 validated hits from the top novel candidates

• 3 days instead of months of screening