A Startup Is Using AI to Speed Up Drug Testing. Here's What That Means.

A company called 10x Science has just raised $4.8 million to build AI tools that help speed up one of the slowest parts of drug development: testing how new drug candidates actually work in the lab.

The company is part of Y Combinator's startup accelerator program. Its focus is on a real problem in the drug industry: computers can now design thousands of potential drug candidates very quickly, but scientists still need to test each one in a physical laboratory to see if it actually works. That testing process is the bottleneck.

Why This Matters Now

Drug companies have recently started using AI and machine learning to dream up new drug candidates. It's like a very smart computer playing with millions of molecular Lego blocks to find combinations that might treat disease. The computers are great at this — they can generate candidates much faster than humans could.

But here's the problem: once you have all these candidates, someone still needs to test them the old-fashioned way. Scientists have to take each drug candidate into a real lab, mix it with human or animal proteins, and measure what happens. This involves specialized equipment, takes weeks or months per drug, and costs a lot of money.

That gap between how many candidates the computers can dream up and how many scientists can actually test is what 10x Science wants to fix.

How 10x Science's Approach Works

The company is building AI systems trained on data about protein structures and how drugs stick to proteins. Think of it like teaching a computer to predict what will happen when two puzzle pieces come together — without having to physically put them together first.

The company says its system can characterize drugs faster and handle many candidates at once, rather than testing them one by one. The details of exactly how it works are still private.

Who Else Is Doing This

Other companies in this space include Recursion Pharmaceuticals, Exscientia, and Atomwise. They've all been working on using AI for drug discovery. Worth flagging: the pharmaceutical industry has seen similar mismatches before when new technology made it easy to generate data faster than the old methods could process it.

10x Science is taking a narrower approach than some competitors. Instead of trying to be a complete drug discovery platform, it's focused specifically on the testing and verification step. This could make it easier for drug companies that already have their own discovery tools to adopt 10x Science's solution.

Funding and Next Steps

The $4.8 million seed round is the company's first major institutional investment beyond Y Combinator's own backing. Yahoo Tech reported the funding, though the other investors involved have not been named.

What Success Looks Like

For 10x Science to win over pharmaceutical companies, the system will need to prove it makes drug testing faster and more accurate than traditional methods. Drug companies are cautious about new technology when it comes to drug development — there are strict regulatory rules, and any shortcuts need to be validated carefully.

In this author's view, the real test will be whether the AI predictions hold up as well in the real lab as they do in the computer model.

Analysis: This is part of a broader shift toward AI-powered tools for drug development infrastructure. The opportunity is real, but it requires sustained investment in both technology and meeting regulatory requirements. 10x Science's focused approach — tackling one specific bottleneck rather than trying to overhaul the entire drug development process — may be exactly what the industry is ready for right now.