20
November
2020

High Performance Computing Helps Researchers Predict Whether a Drug Will Harm Your Heart

Subscribe for more like this

Share

By Ben Fineman

Using high performance computing for reasearch can seem abstract and difficult to visualize. What exactly is being researched, and what problems are being solved? That’s one of the reasons I’m excited about this second webinar in a four- part series with Oracle, where Dr. Igor Vorobyov, a researcher from the University of California, Davis will team up with Rajib Ghosh, global solutions architect for Oracle, to talk about a critical and common medical issue that I was previously unaware of: cardiotoxicity.

Cardiotoxicity. It’s a terrible-sounding word for an equally terrible circumstance: it describes when a drug to cure one ailment also does harm to a patient’s heart. Cardiotoxicity is a serious and expensive problem for the pharmaceutical industry; nearly 10% of drugs in the past four decades have been pulled from the worldwide clinical market due to cardiovascular concerns, despite efforts to gauge cardiotoxicity risk in the development and testing stages. It’s also a serious problem for doctors and patients, who have to balance the need for a cure with the risk of heart problems.

Researchers in the Vorobyov Lab and Clancy Lab at the University of California, Davis, are working with Oracle for Research to develop an in silico (computer) model to better understand and predict the likelihood of pharmaceutical compounds to cause heart pattern disturbances, known as arrhythmias. Led by Igor Vorobyov, PhD, the first task for the team is testing the model against compounds that are known to bind to a cardiac ion channel protein encoded by a specific human gene (hERG), blocking critical molecular interactions that manage the heart’s rhythms. Once the model is tested and trained, it will try to predict the pro-arrythmic proclivities of compounds whose effects are not known.

In this second in a four- part series with Oracle, Igor Vorobyov will team up with Rajib Ghosh, global solutions architect for Oracle, to discuss how high performance computing can accelerate medical research. Register here.

Cardiotoxicity occurs with several classes of drugs, especially when they are administered at higher doses. It is particularly prominent in drugs used to treat cancer. The same drugs that kill cancer cells may also attack the tissues of the heart. Sometimes the effects are immediate. Sometimes they occur years later, after a patient has long been cancer free. Doctors who treat cancer work with cardiologists – doctors who treat heart ailments – to mitigate the cardiac harm that a patient undergoing chemotherapy may suffer. Their challenge is to administer enough of the drug to eliminate the cancer while keeping the dosage low enough to avoid damaging the heart. It would undoubtedly be better if they had better options, including new drugs that can fight cancer without harming the heart, even at high dosages.

The research team also aims to resolve an additional challenge that is not well-addressed by current pharmaceutical testing protocols. Cardiotoxicity often occurs in heart tissue that is unhealthy, but in many patients, unhealthy tissue is masked by healthy heart tissue. To truly assess cardiotoxicity risk, drugs must be tested in the context of comorbidities: what will this drug to do a patient who already has complicating factors that increase his or her risk of heart problems, like a cancer patient who also suffers from diabetes?

The focus of the research project is to use advanced molecular dynamics simulations to develop an AI-driven in silico multi-scale functional model that predicts – in the early stages of drug development – the likelihood that a drug will harm the heart, at what doses that harm will occur, and what additional risks might exist in the context of common comorbidities. The work is computationally intensive, and requires high performance CPU and GPU processors that exceed the researchers’ local computing resources. By moving the research work to Oracle Cloud Infrastructure, Vorobyov and his team gained access to enterprise scale computing including high performance bare metal CPU and GPU shapes that can be used in combination with Oracle’s ultra-fast networks. This enables the team to process more data and run more simulations more quickly.

Introducing high performance, scalable, enterprise cloud computing accelerates discovery and is transforming medical research and treatment. If successful, the work that. Vorobyov and his team are doing will save pharmaceutical companies billions of dollars, which could potentially lower the consumer cost of drugs. More importantly, it will save lives, more accurately predicting safer dosages and supporting faster development of new and different drugs that more effectively fight disease and while leaving untargeted tissues untouched.

Learn more about this critical research using high performance computing. Sign up to attend the webinar on December 8 at 12 p.m. ET, how UC Davis and Oracle for Research are combating drug-induced cardiotoxicity. And, you can replay the first webinar: Igniting Research with Oracle High Performance Computing.