Using Big Data to Find Lifesaving Drugs

A better cure for cancer and other illnesses could already be in existence, hidden right under our noses. The problem is that new potentially lifesaving drugs are created much faster than scientists can study them. Millions of untested compounds wait, jumbled together in vast repositories called compound libraries. 

The current process to screen for possible therapies relies on a gene engineered to mimic a specific target – a gene from a cancer cell, for example – reacting to compounds just as the target would react. Every compound that provokes that reaction is a potential therapeutic. Finding the right compounds, however, is a time-consuming and expensive process.  

At UCSF, data scientists Steven Altschuler, PhD, and Lani Wu, PhD, have developed a game-changer: a faster, less expensive screening approach. By using computational analysis and image recognition to test for multiple targets simultaneously, this method also can digitally identify many compounds that generate the desired response. Later, if another researcher looks for a compound that produces the same response, the screening process doesn’t have to start from scratch. 

Equally important, the cost for each test drops from hundreds of dollars to a dollar or less, Altschuler says. 

Altschuler and Wu, both professors in the Department of Pharmaceutical Chemistry, are a husband-and-wife team whose careers have moved in tandem for 30 years. Before entering academic medical research, they were together at Microsoft and at the biotech firm Rosetta Inpharmatics (now owned by Merck). It was that combination of skill sets and some out-of-the-box thinking that led them to the notion of using image-recognition software on a cellular level. 

The marriage of big data and biology is entering a golden era, and the imaginative and collaborative approach of UCSF’s Wu and Altschuler is helping lead that effort.