How AI simplifies data management for drug discovery


MIT Technology Review

Calithera conducts registered clinical studies on its products to assess their safety, whether they are effective in patients with certain gene mutations, and how well they work in combination with other therapies. The company needs to collect detailed data from hundreds of patients. While some of the studies are still in the early stages and involve only a small number of patients, others involve more than 100 research centers around the world.

“One of the biggest challenges in the life sciences world is the vast amount of data we generate, more than any other business,” said Behrooz Najafi, Calithera’s leading IT strategist. (Najafi is also the chief information and technology officer for healthcare technology company Innovio.) Calithera needs to store and manage the data while ensuring that it is available years from now, if necessary. It must also meet the specific FDA requirements regarding the generation, storage and use of the data.

Even something seemingly as simple as updating a file server must follow a strictly defined FDA protocol with multiple testing and verification steps. Najafi says all of these compliance-related data battles can add 30 to 40% to the overhead costs of a company like his, both in terms of direct costs and employee hours. These are resources that could otherwise be used for more research or other value-adding activities.

Calithera has bypassed much of this additional cost and has significantly improved its ability to track its data by placing it in a secure “storage container,” as Najafi calls it, a restricted area for regulated content, part of a larger cloud document management application has made mainly by artificial intelligence. AI never sleeps, never gets bored and can learn to differentiate between hundreds of different types of documents and forms of data.

Here’s how it works: Clinic or patient data is entered into the system and scanned by AI, which detects specific characteristics such as accuracy, completeness, regulatory compliance and other aspects of the data. AI can report when a test result is missing or a patient has failed to submit a required diary entry. It knows who can access certain types of data and what they can and cannot do with it. It can detect and ward off ransomware attacks. And it can automatically document all of this to the satisfaction of the FDA or other regulatory agency.

“This approach relieves us of the compliance burden,” says Najafi. Once the data from its many research sites is on the platform, Calithera knows that the AI ​​is making sure it’s safe, complete, and regulatory compliant, and everyone is reporting issues.

As Najafi notes, managing drug discovery data to meet research and regulatory requirements can be tedious and expensive. The life science industry can borrow data management techniques and platforms developed for other industries, but they need to be modified to handle the level of security, validation and detailed audit trails that are a way of life for drug developers. AI can streamline these tasks and improve the security, consistency, and validity of data – freeing up overhead for pharmaceutical companies and research organizations to do their core business.

A complex data management environment

Regulatory compliance helps ensure that new drugs and devices are safe and function as intended. It also protects the privacy and personal information of thousands of patients participating in clinical trials and post-market research. Regardless of their size – giant global corporations or tiny startups trying to bring a single product to market – drug developers must follow the same standard practices to document every scrap of information related to a clinical trial, to review validate and protect.

When researchers conduct a double-blind study, the gold standard for demonstrating a drug’s effectiveness, they must keep patient information anonymous. But they have to slightly deanonymize the data later to make it identifiable so that patients in the control group can receive the test drug and the company can – sometimes for years – keep track of how the product behaves in real use.

The data management burden is heavy on emerging and midsize life science companies, says Ramin Farassat, chief strategy and product officer at Egnyte, a Silicon Valley software company that makes and supports the AI-enabled data management platform used by Calithera and hundreds of other companies. scientific company.

“This approach relieves us of the compliance burden,” says Najafi. Once data from its many research sites is on the platform, Calithera knows that the AI ​​is making sure it is safe, complete, and compliant with all regulations and reporting any issues.

Download the full report.

This content was created by Insights, the custom content division of MIT Technology Review. It was not written by the editorial staff of the MIT Technology Review.

Source Link

Leave a Reply