AWS launches Bio Discovery

Amazon Web Services on April 14 launched Amazon Bio Discovery, a new artificial intelligence tool for early drug research that scientists can use without coding. (reuters.com) Early-stage drug discovery is the search for molecules worth testing in a lab, a process that often starts with huge lists of possible candidates and many rounds of elimination. Amazon said Bio Discovery lets researchers use biological foundation models — artificial intelligence systems trained on large biology datasets — to generate and rank those candidates. (aboutamazon.com) Amazon said the product includes an artificial intelligence agent that helps scientists choose models, set inputs and review results in plain language. The system also connects to outside lab partners so promising candidates can be synthesized and tested, with the data fed back into the next round of design. (reuters.com) (aws.amazon.com) The pitch is speed and fewer handoffs between software teams and lab teams. Rajiv Chopra, vice president of healthcare artificial intelligence and life sciences at Amazon Web Services, told Reuters that a task that once took 18 months to produce 300 potential drug candidates can now produce 300 candidates in a couple of weeks. (reuters.com) Amazon is entering a crowded field where drugmakers and technology companies have been racing to use artificial intelligence to shorten the earliest, most failure-prone part of drug development. Reuters said the launch adds to competition to make computational drug research faster and easier for scientists who are not machine-learning specialists. (reuters.com) (axios.com) Amazon’s argument is that the bottleneck is no longer just computing power. In its launch materials, the company said many bench scientists lack direct access to these tools, while computational biologists are stuck evaluating fast-changing models and stitching together disconnected workflows. (aws.amazon.com) (aboutamazon.com) Amazon said Bio Discovery starts with more than 40 biology models and lets customers add their own or licensed third-party models. The company said computational biologists can build reusable no-code workflows, while bench scientists can run variations of experiments in the same system. (aws.amazon.com) Amazon named Bayer, the Broad Institute and Voyager Therapeutics as early adopters. It also said 19 of the top 20 global pharmaceutical companies already use Amazon Web Services cloud products, giving the company an installed base to sell into. (reuters.com) In one example, Amazon said work with Memorial Sloan Kettering Cancer Center used multiple models to generate nearly 300,000 new antibody molecules and narrow them to 100,000 candidates for lab testing by Twist Bioscience. Amazon said that reduced work that can take months to a period of weeks. (reuters.com) (aboutamazon.com) Amazon told Reuters the service is meant to support scientists and contract research organizations, not replace them. The company said it will start with a free trial that includes five experimental units before moving to subscription pricing. (reuters.com)