SAN FRANCISCO, Jan 13 (Reuters) - AstraZeneca has agreed to buy Boston-based Modella AI, the companies said on Tuesday, as the drug industry ​increases its use of artificial intelligence to accelerate the discovery of ‌new drugs.

The companies did not disclose financial terms. In a press release, Modella AI said its "foundation ‌models" and AI agents would be integrated into oncology research and development to support clinical development and biomarker discovery.

"Oncology drug development is becoming more complex, more data-rich and more time-sensitive," said Gabi Raia, Modella AI's chief commercial officer, adding that ⁠joining AstraZeneca would allow ‌it to deploy its tools in global trials and clinical settings.

AstraZeneca said that this was the first acquisition of an ‍AI firm by a big pharmaceutical company.

In an interview at the JP Morgan Healthcare Conference, AstraZeneca Chief Financial Officer Aradhana Sarin said the acquisition would "supercharge" the company's quantitative pathology ​and biomarker discovery efforts by bringing more data and AI capabilities in-house.

The ‌deal was one of a number of pacts between major drug firms and AI companies that were unveiled at the healthcare conference, including a $1 billion collaboration between Nvidia and Eli Lilly. They plan to build a new research lab using Nvidia's latest-generation AI chips.

Modella will accelerate AstraZeneca’s efforts to make pathology ⁠more quantitative - using computers to analyze biopsies for ​relevant proteins and correlate them with clinical ​data - so AstraZeneca can develop “highly targeted biomarkers and then highly targeted therapeutics,” Sarin said.

The deal is an expansion of a multi-year collaboration ‍that the companies ⁠unveiled in July.

Sarin said that partnership served as a "test drive," adding that AstraZeneca ultimately wanted Modella's data, foundation models and AI talent in-house.

She ⁠said AI tools could be used to more rapidly select patients for drug trials, which could ‌increase the odds of clinical success and cut related costs.

(Reporting ‌by Maggie Fick; Editing by Thomas Derpinghaus)