IBM AI will help researchers find a drug to treat coronavirus. Another tool based on machine learning will allow scientists to quickly look for reliable data among scientific papers.
IBM noted that they would like to help researchers better understand the nature of COVID-19 and how to treat it. To do this, they suggest using the power of two AIs. One of them can use data from thousands of authorized sources and aggregate information from them. This will help researchers quickly look for potential treatments and provide new information to healthcare professionals.
Skilled researchers can ask AI questions. Using machine learning, he quickly extracts information from 13,335 documents, the list of which is increasing daily. IBM noted that AI is capable of adding up to 100 thousand PDF files per day.
The second tool is a “molecular researcher.” It has a graphical interface that can be used to filter therapeutic agents according to characteristics of interest, search for similar molecules and visualize their physical structure. Scientists can also export data from this tool to other applications for further analysis.
“In order to systematize the identification of new treatments for COVID-19, we also provide free access to the functional genomics platform for the pandemic period”, wrote Dario Gilles, head of IBM research on the blog. “Designed to detect the molecular characteristics of 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 their connections”.