Accelerating drug discovery with AI and chemistry
Principal Scientist, Johnson & Johnson Innovative Medicine
Rebecca Alford (MCS 2016) loves watermelon and she approaches picking the perfect one with a scientist鈥檚 eye. It鈥檚 a careful process of observation and inference, not unlike how she uncovers promising directions for new medicines as a principal scientist specializing in structure-based drug design at . 听
At J&J, Rebecca develops computational methods to design complex molecules, including peptides and RNAs. Her work sits at the intersection of chemistry, physics and data science, using computer models and AI to explore what is possible long before a molecule is ever made in the lab. She then works closely with chemists and biologists to test these molecules.
鈥淚f you know what features to look for in a watermelon, like the coloring, you know it will be great,鈥� Rebecca says. 鈥淚 use the same idea in my work, identifying features of molecules that increase their chances of becoming a successful drug.鈥�
Sounds easy, but the number of possible molecules Rebecca has to search is 10 to the 60th power 鈥� an almost unimaginable quantity to explore. If you had a pile of watermelons that size, the length would measure in thousands of light years. The sheer volume of options makes drug discovery time consuming, expensive and fraught with dead ends.
鈥淒eveloping a drug the existing way takes at least 10 years and costs billions of dollars. By infusing algorithms with chemistry, physics and AI, I鈥檓 accelerating the search process,鈥� she says.
Rebecca鈥檚 work is also deeply personal. Motivated in part by her own experience living with a rare genetic disorder, she is driven by the urgency of patients waiting for new treatments.
鈥淲e do this work because there are people suffering and we are trying to deliver better therapies,鈥� she says. 鈥淭he problem is so big and so hard, but we鈥檝e reached the point where modeling and AI can help get medicines to patients faster. That鈥檚 what keeps me going.鈥�
Story by Elizabeth Speed