IBM scientists secure patent on machine learning models for drug discovery

The models will predict therapeutic indications and side effects from several drug information sources

The invention of the new machine learning models is likely to pave way for faster discovery of more effective and safer medication, said IBM.

In this regard, a cognitive association engine was deployed by IBM to detect significant relationships between predicted therapeutic indications and side effects.

The engine also features a visual analytics system to back the interactive exploration of these associations.

According to IBM, the novel method could enable researchers in pharmaceutical firms to create hypotheses needed for drug discovery.

The technology giant said that strongly correlated disease-side-effect pairs identified by the cognitive association engine could benefit drug discovery in several ways.

The company said if a new drug is being developed for a disease that is strongly correlated with severe side effects, then special care could be given for moderating its formulation and dosing during clinical studies to avoid serious safety problems.

IBM Research Center for Computational Health senior manager and program director Jianying Hu said:  “As inventors at IBM, we have the opportunity to help solve real-world problems.

“Our team is dedicated to this research and we continue to search for new ways to improve people’s health around the world through innovation and invention.”

Two of the main reasons as per IBM that make a drug falter in a clinical study are lack of efficacy and harmful side effects with each making up for about 30% of failures.

To address the issues, IBM stated that useful insights derived through computational models and machine learning methods from vast amounts of data on drugs and diseases from different sources hold great potential for bringing down attrition rates and accelerating the drug discovery process.

Image: IBM Research staff member Ping Zhang (left) and Program Director, Center for Computational Health Jianying Hu (right) with their newly patented invention to help drug discovery researchers identify which drug indications are typically linked to which side effects. Photo: courtesy of IBM.