Ai-pdp

For each colorectal cancer patient, Vivan Therapeutics will build 500,000 fruit fly avatars genetically matched to that person’s tumour, then screen thousands of drug combinations to find the most effective treatment. This matters because colorectal cancer is genetically diverse—nearly 80% of patients can be assigned to one of 128 distinct genetic signatures, yet most treatments are one-size-fits-all. Current approaches often fail because they do not account for a tumour’s specific mutations, and finding effective drug combinations through conventional clinical trials is slow and expensive. Vivan’s platform is the only one that allows personalised, high-throughput drug screening in a living organism. By running these screens across all 128 genetic profiles, the company will generate a massive dataset of which drug combinations work for which tumour type. That data will train machine-learning algorithms to create an AI tool that can predict effective, affordable treatments—often using non-cancer drugs that are less toxic—without needing to build fly avatars for every future patient. If successful, this could shift colorectal cancer treatment from trial-and-error prescribing to rapid, data-driven precision medicine, reducing both side effects and costs.