Nvidia-backed AI startup SandboxAQ creates new data to speed up drug discovery

FILE PHOTO: A NVIDIA logo is shown at SIGGRAPH 2017 in Los Angeles, California, U.S. July 31, 2017. REUTERS/Mike Blake
SAN FRANCISCO (Reuters) -SandboxAQ, an artificial intelligence startup spun out of Alphabet's Google and backed by Nvidia, on Wednesday released a trove of data it hopes will speed up the discovery of new medical treatments by helping scientists understand how drugs stick to proteins.
The goal is to help scientists predict whether a drug will bind to its target in the human body.
But while the data is backed up by real-world scientific experiments, it did not come from a lab. Instead, SandboxAQ, which has raised nearly $1 billion in venture capital, generated the data using Nvidia's chips and will feed it back into AI models that it hopes scientists can use to rapidly predict whether a small-molecule pharmaceutical will bind to the protein that researchers are targeting, a key question that must be answered before a drug candidate can move forward.
For example, if a drug is meant to inhibit a biological process like the progression of a disease, scientists can use the tool to predict whether the drug molecule is likely to bind to the proteins involved in that process.
The approach is an emerging field that combines traditional scientific computing techniques with advancements in AI. In many fields, scientists have long had equations that can precisely predict how atoms combine into molecules.
But even for relatively small three-dimensional pharmaceutical molecules, the potential combinations become far too vast to calculate manually, even with today's fastest computers. So SandboxAQ's approach was to use existing experimental data to calculate about 5.2 million new, "synthetic" three-dimensional molecules - molecules that haven't been observed in the real world, but were calculated with equations based on real-world data.
That synthetic data, which SandboxAQ is releasing publicly, can be used to train AI models that can predict whether a new drug molecule is likely to stick to the protein researchers are targeting in a fraction of the time it would take to calculate it manually, while retaining accuracy. SandboxAQ will charge money for its own AI models developed with the data, which it hopes will get results that rival running lab experiments, but virtually.
"This is a long-standing problem in biology that we've all, as an industry, been trying to solve for," Nadia Harhen, general manager of AI simulation at SandboxAQ, told Reuters on Tuesday. "All of these computationally generated structures are tagged to a ground-truth experimental data, and so when you pick this data set and you train models, you can actually use the synthetic data in a way that's never been done before."
(Reporting by Stephen Nellis; Editing by Leslie Adler)

Try Our AI Features

Explore what Daily8 AI can do for you:

Learn with AIFact CheckingText to SpeechAI Summary

Related Articles

The Star

7 hours ago

• The Star

Google offers to tweak search results to promote rivals, stave off EU antitrust fine, documents show

FILE PHOTO: A logo is pictured at Google's European Engineering Center in Zurich, Switzerland July 19, 2018. Picture taken July 19, 2018. REUTERS/Arnd Wiegmann/File Photo BRUSSELS (Reuters) -Alphabet's Google has proposed more changes to its search results to better showcase rivals in a bid to stave off a possible hefty EU antitrust fine, according to documents seen by Reuters. Google's latest proposal came three months after the European Commission charged the U.S. tech giant with favouring its own services such as Google Shopping, Google Hotels and Google Flights over rivals in breach of the Digital Markets Act (DMA). The landmark DMA sets out a list of dos and don'ts for Big Tech aimed at reining in their power and giving rivals more room to compete and consumers more choices. Under Google's new proposal a vertical search service (VSS) selected on objective and non-discriminatory criteria would get its own box at the top of the search page with the same format, information and features as Google's, the document said. The box would contain three direct links picked by the VSS, to hotels, airlines, restaurants and transport. Other VSS, which are specialised search engines within Google, would be ranked below but without a box unless users click on them. "We do not agree with the (Commission's) preliminary findings' position but, on a without prejudice basis, we want to find a workable solution to resolve the present proceedings," the documents sent by both Google and the Commission to the rivals said. The rivals will provide feedback at a July 8 meeting called by the Commission. A number of rivals, who did not want to be named ahead of the meeting, told Reuters that the changes still do not go far enough to ensure a level playing field. (Reporting by Foo Yun Chee; Editing by Susan Fenton)

The Star

8 hours ago

• The Star

Musk's xAi increases yield offer on $5 billion debt raise, source says

FILE PHOTO: A 3D-printed miniature model of Elon Musk and the xAI logo are seen in this illustration taken January 23, 2025. REUTERS/Dado Ruvic/Illustration/File Photo NEW YORK (Reuters) -Elon Musk's xAi is increasing the yield it is offering on a $5 billion debt raise led by Morgan Stanley, a source with knowledge of the matter said on Friday. xAi is offering to pay 12.5% yield on $3 billion in bonds, said the source who asked for anonymity to disclose non-public information. Previously, sources told Reuters the company had offered a 12% yield. xAi is also offering 12.5% fixed yield on a $1 billion term loan and set to price a $1 billion term loan B at 725 basis points over the Secured Overnight Financing Rate, known as SOFR. The term loan B is set to be priced at a discount of 96 cents on the dollar, the source said. The initial offering on the securities was 12% on the fixed loan and 700 basis points over the SOFR on the floating rate loan. Junk-rated bonds paid an average yield to maturity of 7.602%, according to ICE BofA High Yield Index. The deadline for investor commitments was extended from Tuesday to Friday and allocations will be done one day after closing, the source said. If the deal closes on Friday, allocations will happen on Monday. An increase in the yield offer could mean that investors had probably agreed to buy the debt only for a higher yield. The borrower also has lesser flexibility on pricing when investor demand is modest. The xAI offering, which was reported on June 2 as Musk and U.S. President Donald Trump traded barbs over social media, did not receive overwhelming interest from high-yield and leveraged loan investors, Reuters reported earlier this week. Unlike Musk's debt deal when he acquired Twitter, Morgan Stanley did not guarantee how much it would sell or commit its own capital to the deal, in what is called a "best efforts" transaction, according to one person familiar with the terms. xAi did not immediately respond to a request for comment. Morgan Stanley declined to comment. (Reporting by Tatiana Bautzer; Additional reporting by Matt Tracy; Editing by Mark Porter, Alexandra Hudson)

The Star

13 hours ago

• The Star

Norway plans temporary ban on power-intensive cryptocurrency mining

FILE PHOTO: Representations of cryptocurrencies are seen in this illustration, August 10, 2022. REUTERS/Dado Ruvic/Illustration/ File Photo COPENHAGEN (Reuters) -Norway aims to impose a temporary ban on the establishment of new data centres that mine cryptocurrency with the most power-intensive technology, in order to conserve electricity for other industries, the Nordic country's government said on Friday. "The Labour Party government has a clear intention to limit the mining of cryptocurrency in Norway as much as possible," Minister for Digitalization and Public Administration Karianne Tung said in a statement. "Cryptocurrency mining is very power-intensive and generates little in the way of jobs and income for the local community," she added. A temporary ban could be introduced during the autumn of 2025, the government said. (Reporting by Louise Breusch Rasmussen, editing by Terje Solsvik)