Obstructions to data access are preventing solutions to humanity's most pressing problems. Researchers publish thousands of scientific papers every day. Healthcare professionals create hundreds of millions of medical records. People and organisations around the world are collecting vast amounts of data. At BenevolentAI we use our technology to empower our scientists to make sense of it all. No patient should suffer without treatment. But thousands of diseases today have no cure. Access to more data increases our chances of discovering new and better medicines. This rings true for everyone working in drug discovery.
We leapt into action for the Coronavirus disease (COVID-19) pandemic. Without a vaccine available any time soon we need medicines that can prevent life-threatening complications. Scientists used our AI platform to find existing drugs that could potentially treat patients. Our research identified the rheumatoid arthritis drug Baricitinib in 48 hours. In February we openly published our initial findings, and later research. Eli Lilly is the owner of Baricitinib. They have now initiated large scale clinical trials. We expect results within the next few months. This end-to-end pace is unheard of in pharma.
All researchers should be sharing findings related to this outbreak rapidly and openly. Many have done so already. We need this to inform the public health response and help save lives. That's why we joined many others signing the Wellcome Trust pledge, committing to sharing our research findings in the public domain. Many other data holders have openly shared information during the pandemic. Academic literature publishers have removed paywalls for related articles. The Copyright Clearance Centre is maintaining a list of these free resources. The Allan Institute for AI released an Open Research Dataset of relevant literature. EMBL-EBI and partners created a COVID-19 Data Portal of biomedical data. There are many more. This unprecedented availability of open data has had clear benefits for the research already undertaken in the short period since the pandemic hit.
The global health crisis provides a unique set of circumstances. But it allowed us to see what happens if people share more data. Data is necessary for drug discovery. Access to more data increases the chance of developing medicines that are more efficacious in the patients in which they are prescribed for. It also increases the chance of avoiding pursuit of the wrong medicines. Developing a new drug can take a decade or more, and most drug programmes fail. This means it costs billions to bring a new drug to market. Many disease areas are not commercially viable areas of research. Access to more data will not only bring down the price of drugs, it will also make it cost effective for rare diseases that are ignored today.
There are good reasons for not opening data. Collecting, cleansing, evaluating, aggregating, formatting, and even sharing data takes effort. We need to fairly compensate those undertaking this work. Patient data has privacy and consent implications. Information security is important. But none of these challenges are insurmountable. Some data isn't shared simply because people don't know about it. Some data isn't shared because owners don't know what license to use. Some data isn't shared because of usage restrictions the owner didn't intend. We need to encourage people to share data while protecting the investment made in creating and curating it.
If you have data, share as much of it as you can. Make its existence more visible. Get the right consent from patients and other data subjects. Use latest data licenses designed for sharing in today's machine learning world. We need to make this easy so we can all benefit.
VP Intellectual Property