Coronavirus particles

Global coronavirus research collated into AI-powered database

Image credit: Science Photo Library

A database collating all of the world’s research on coronavirus has been created by scientists in the hope that this could help accelerate the search for drugs to combat the disease.

A team at The Institute of Cancer Research (ICR) in London has fast-tracked development of the powerful database by adapting an existing system called canSAR, which already pulls together data from across cancer research and drug discovery.

The new resource will make vast amounts of data on the biology and treatment of Covid-19 freely available.

The new ‘knowledgebase’ draws in data published across the world on viral proteins; interactions of viral proteins with human proteins; drugs and drug mechanisms, and clinical trials. 

It is believed to be the world’s first one-stop shop for research on Covid-19 and related diseases such as SARS and MERS.

Coronavirus-CanSAR has been developed by a team of researchers funded by Cancer Research UK and the Wellcome Trust and is the first phase of a push to capture the rapidly changing information on coronaviruses. It should help researchers rapidly repurpose drugs from across the whole of medicine, including cancer.

The online knowledgebase will be updated weekly and pulls in every 3D structure of all biological molecules published in the international Protein Databank, totalling over 450,000 so far. This information includes more than 830 3D structure snapshots of 47 proteins made by the novel coronavirus, MERS and SARS.

Using canSAR’s AI technology, the ICR team analysed more than four million sites across all proteins in its database, including more than 8,000 from coronavirus protein structures.

The researchers identified several ‘druggable cavities’, suitable sites for novel antiviral drugs targeted at the contact points between human and virus proteins.

Coronavirus-CanSAR uses AI to generate complex three-dimensional maps of how virus proteins interact with human proteins that look like visualisations of space. These maps are based on data from almost a million interactions, highlighting those that could potentially be disrupted with new drugs.

The team will also assess the quality of data published on how viral proteins interact with human proteins or existing drugs, so that the system can point users to the most promising leads for new treatments.

Project leader professor Bissan Al-Lazikani, ICR head of data science, said: “We are launching the initial phase of the world’s first one-stop shop for coronavirus research, designed to rapidly pull together data on Covid-19 as it becomes available.

“It is more than just a database. It’s an intelligent system which visualises vast amounts of complex data in a way that is easy for researchers to understand and act upon; predicts which approaches to treatment are most likely to work, and objectively assesses the information emerging from clinical trials.

“We hope that Coronavirus-CanSAR can play a big role in the search for new drug treatments for Covid-19 and can become a vital tool for future research into viral diseases, too. Unfortunately, nothing like this existed during the past SARS and MERS outbreaks and we need to ensure that it is maintained and developed as part of our rapid response to future pandemics.

“The faster we beat Covid-19, the faster we can get back to defeating cancer, and as we’ve demonstrated in this work, one field can benefit from knowledge generated by another. We will take the lessons from Coronavirus-CanSAR back to benefit our cancer research.”

Sign up to the E&T News e-mail to get great stories like this delivered to your inbox every day.

Recent articles