Claude AI Successfully Designs Functional Proteins in Major Scientific Breakthrough

Anthropic says Claude designed working protein binders

The company reported the results itself, tested by two outside labs, and kept the work off its most capable model. Anthropic says its Claude models designed working protein binders and ran a chemical-analysis job in minutes. It published the two experiments on Tuesday. The company reported the results itself. It framed them as early evidence that Claude can speed up parts of drug development. Anthropic shared the work on its research blog. The headline numbers are Anthropic's own. Claude designed protein binders against 15 targets and succeeded against 14, the company said. Between 22 and 35 percent of its designs bound to their target, depending on the setup. Anthropic says 10 to 15 percent is typical in the field today. Some of the strongest designs bound several times more tightly than the best previously published result, it added. The wet-lab checks were not done in-house. Two outside firms, Adaptyv Bio and Twist Bioscience, independently produced and tested Claude's designs, Anthropic said. That physical validation is the part of protein work that still takes weeks, whatever the software. How Claude designed the binders A binder is a small protein built to latch tightly onto a target, which is how many modern drugs work. Designing a new one from scratch, known as de novo design, has historically taken protein engineers months per target. The work overlaps with the wider field of AI drug discovery. Anthropic used two models, Opus 4.8 and a preview of its Mythos model. It ran them inside Claude Science, its research workbench. The models got a long prompt, internet access, connectors and a large pool of GPUs. The company said it then left the models to work autonomously. Claude chose where on each target to bind, the company said. It orchestrated existing open-source design and folding models, then screened the candidates. The compute was substantial. Anthropic gave the models up to 12,500 Nvidia H100 hours over a 48-hour session in one mode. A second mode used up to 2,500 H100 hours per target. Designing against one target at a time worked best, it said. The Mythos preview hit a 35.1 percent success rate that way. In all, the campaign produced 354 confirmed binders from 1,320 designs. Where it did well, and where it failed Some results stood out. Against a target called RBX1, the Mythos preview hit a 40 percent success rate, Anthropic said. Human entrants in an Adaptyv Bio competition managed 3.7 percent. Claude's top design beat the winning entry. Then there was TNFα, the protein that anti-inflammatory drugs such as Humira block. Opus 4.8 designed binders that worked across human, monkey and mouse versions. Oddly, the Mythos preview failed on that target. Anthropic said it was not sure why the less capable model succeeded where the stronger one did not. It did not work everywhere. Against maltose-binding protein, a notoriously smooth target, none of 90 designs was confirmed to bind, Anthropic said. It managed only modest results against BBF-14, a synthetic protein used as a hard benchmark. The company said it plans further testing to confirm its hit rates and binding measurements. It has released its prompts and data. The chemistry test used a public model The second experiment used Claude Opus 5, the most capable model Anthropic offers to the general public. The task was chemical analysis, the routine work of checking what a compound is and how pure it is. Chemists normally do this with two techniques, NMR and LC-MS, and the slow part is reading the raw files each instrument spits out. Anthropic gave Opus 5 only the raw files from a contract lab and a two-sentence prompt, with no vendor software and no operator. Working inside Claude Science, the model returned processed results in 23 and 19 minutes, running the two analyses in parallel, the company said. Its numbers matched the lab's. Purity came out at 96.4 percent against the lab's 96.33 percent, and hydrogen counts were within a rounding margin. Two details are worth noting. Claude caught and corrected its own error, Anthropic said, after a first pass overstated how many peaks had shifted in a follow-up reading. And it proposed the same confirmatory test the lab had independently run three days earlier. For the LC-MS file, which came in an undocumented vendor format, the model worked out the encoding and checked its reading against the instrument's own totals before analysing anything, the company said. The comparison Anthropic draws is with time. A chemist doing this by hand takes roughly 30 minutes to an hour per sample, and the lab's finished report for this one arrived four days after the first measurement. Claude produced its report inside 25 minutes. The results are self-reported, and dual-use The findings come from Anthropic, about its own models. They have not been through outside peer review, though the two external labs tested the protein binders. The company said it assessed its models "holistically" and would publish more characterisation to confirm the figures. Anthropic also flagged the risk in its own work. Autonomous biological design is dual-use, it said: the same capability that speeds up medicine could help a bad actor build a bioweapon. For that reason, protein design and other sensitive biology tasks remain blocked on Claude Fable 5, its most capable model. It is building a vetted access program for scientists instead. The company has run its own tests on how far its bioweapon filters hold. The context matters for how much to read into the results. Anthropic's own post notes that AI has moved fastest in fields such as maths, where an answer can be checked quickly. It has moved more slowly in the life sciences, where verification is expensive. It casts this work as foundational, and says it wants Claude to eventually run drug development end to end. A designed binder, it acknowledged, is only the first step toward an actual drug.

Claude Can Now Design Proteins -- And Scientists Tested It in the Lab

Anthropic's Claude is moving deeper into scientific research, with the AI model taking on tasks that have traditionally required scientists to spend hours or even weeks working through complex data. In a blog post, Anthropic said Claude was able to design proteins from scratch and analyze chemistry data with limited human input. The company said the experiments show how AI could speed up parts of scientific research and reduce the amount of specialized expertise needed to get work done. In one experiment, Anthropic asked Claude to design small proteins that could attach to specific targets. The company tested the system against 15 targets and found that it successfully produced working designs for 14 of them. Claude generated 1,320 designs during the campaign, ultimately producing 354 working protein binders. Its overall success rate ranged from 22.6% to 35.1%, compared with a typical success rate of about 10% to 15% for protein-design campaigns, according to Anthropic. The company said Claude carried out much of the process on its own. It selected where to design on each target, generated potential designs, ran them through other specialized models and screened the results before the designs were sent to outside laboratories for testing. Markets Anthropic's Pre-IPO Credit Line Is Growing -- and Banks Are Angling for Underwriting Roles Anthropic's revolving credit facility is set to exceed its approximately $10 billion target, as the AI company looks to go public. 2 min read Read this article In one example, Claude achieved a 40% success rate when designing against a protein called RBX1, compared with a 3.7% success rate among participants in a previous protein-design competition. Latest Private Market Opportunities Join 400,000+ Investors However, the model was not successful across the board, Claude struggled with some targets, including maltose-binding protein, where none of its 90 designs were confirmed to work. Anthropic also tested Claude on a more routine scientific task: analyzing chemistry data. The company gave Claude Opus 5 raw files from two common laboratory instruments and asked it to process the information. Claude returned its analysis in 23 minutes for one set of data and 19 minutes for the other, without using the specialized software normally required for the task. The results closely matched the laboratory's own analysis. Claude measured the purity of a sample at 96.4%, compared with 96.33% in the lab's results. Anthropic said this type of work can normally involve manual analysis and significant delays before a finished report is available. In the test, Claude processed both sets of data in parallel and produced a written report within about 25 minutes. The results also highlight the limits and risks of increasingly capable AI systems, biological research capabilities can have both beneficial and harmful uses, including the potential to enable dangerous research. As a result, some of its most advanced biology capabilities remain unavailable for general access, the company noted. Still, Anthropic sees the experiments as an early step toward using Claude across the drug-development process. The company said its broader goal is to have Claude eventually help with the development process from beginning to end. Anthropic CEO Dario Amodei previously remarked that AI could help cure most human diseases within five to 10 years, while arguing that the industry must deliver real breakthroughs in medicine and biology rather than rely on optimistic messaging to win public trust. Amodei made the comments in a post on X, responding to criticism that his public messaging around AI has been disproportionately focused on the technology's risks. He said he views his messaging as roughly balanced between AI's benefits and dangers, citing his writing on the technology's potential to transform health and biology. Photo Courtesy: Stockinq on Shutterstock.com Markets OpenAI Hits the Brakes on Frontier AI Training Over Cybersecurity Fears OpenAI pauses frontier-model development for two weeks after signals indicate it could reach "Critical" cybersecurity capability. 3 min read Read this article This content was partially produced with the help of AI tools and was reviewed and published by Benzinga editors. Market News and Data brought to you by Benzinga APIs To add Benzinga News as your preferred source on Google, click here.

Twist Bioscience benefits from Claude's advances in protein design

The technology group asked Claude Opus 4.8 and its experimental Mythos Preview model to design 'minibinders', small proteins intended to attach precisely to a target protein in order, for example, to block or alter its function. Of 15 targets tested, functional proteins were obtained for 14, with a success rate of 22% to 35% depending on the methods used, versus roughly 10% to 15% generally seen in this type of campaign. For four targets, the best designed proteins matched or exceeded previously published affinity levels. The experiment above all highlights the complementarity between AI models and synthetic biology. Claude can generate, compare and quickly select a large number of sequences, but these then have to be physically produced and tested in the lab. Twist Bioscience and Adaptyv Bio handled precisely this independent validation step. Listed on the Nasdaq, Twist Bioscience specializes in manufacturing synthetic DNA, notably for sequencing and drug discovery. Adaptyv Bio, for its part, operates an automated platform that can rapidly produce and test digitally designed proteins. In Anthropic's experiment, the two companies therefore provided the lab validation of the proteins generated by Claude. This development could gradually strengthen demand for this type of synthetic biology infrastructure if pharmaceutical labs increase their use of AI to generate new candidates. The market quickly reacted to that prospect, with Twist Bioscience shares jumping about 17% Wednesday, toward their highest level since 2021. The stock has now more than quadrupled since the start of the year. Anthropic is also preparing for its arrival in the financial markets. The group confidentially filed paperwork in June for an initial public offering and is now working on its future governance structure, notably on maintaining enhanced voting power for its founders. The deal could become one of the largest technology IPOs ever, with Anthropic already valued at $965bn in its most recent funding round in May.