IBM builds two AI-powered tools for coronavirus research
IBM Corp. today launched two artificial intelligence-powered services that it developed specifically to help advance the study of COVID-19 and potential treatments.
The services were announced by IBM research head Dario Gil in a blog post. The first is a cloud-based molecular explorer that enables researchers to study COVID-19 therapeutic candidates, or molecules that have the potential to form the basis of future treatments.
IBM is launching the tool with a corpus of 3,000 potential therapeutic candidates that it discovered using AI. The company is inviting other organizations to contribute additional molecules to the database now that it’s open to the outside world.
The molecular explorer provides a graphical interface that researchers can use to filter therapeutic candidates based on attributes of interest, find similar molecules and visualize their physical structure. They may also export data from the tool to other applications for further analysis.
“To streamline efforts to identify new treatments for COVID-19, we are also making the IBM Functional Genomics Platform available for free for the duration of the pandemic,” Gil wrote in his post. “Built to discover the molecular features in viral and bacterial genomes, this cloud-based repository and research tool includes genes, proteins and other molecular targets from sequenced viral and bacterial organisms in one place with connections.”
The other service IBM unveiled today, Deep Search, is a search engine researchers can use to find useful data in the large number of academic papers being published about the coronavirus. The tool is aimed at addressing the fact that “as with any large volume of disparate data sources, it is difficult to efficiently aggregate and analyze that data in ways that can yield scientific insights,” Gil wrote.
Deep Search is powered by an AI model that IBM says was trained on thousands of scientific papers. On launch, the service enables researchers to look for data in more than 13,000 studies sourced from the CORD-19 dataset co-created by Microsoft Corp. last month, as well as Clinicaltrials.gov and the DrugBank and GenBank scientific databases.
IBM’s new tools may find use among the scientists who will be working with the COVID-19 High Performance Computing Consortium. The group, which IBM helped launch two weeks ago, includes tech firms, universities and government agencies that will be making supercomputers available for researchers studying the coronavirus. The participating organizations so far have committed 330 petaflops of computing capacity from 16 systems.