The Computational Biology Department Sponsored Recent STEM Junction Symposium
The Computational Biology Department recently sponsored Fox Chapel Area High School’s STEM Junction Symposium, an all-day bioinformatics event hosting 180 students from the Pittsburgh area. The Symposium was organized by Fox Chapel senior Shania Khatri to introduce high school students with an interest in computer science and biology to the growing field of computational biology. The Symposium is one of the first of its kind for a high school audience, but this did not dissuade Khatri. “I was convinced,’ Khatri says, ‘that there should be opportunities for high school students to explore bioinformatics.”
The students in attendance were able to choose from multiple breakout sessions led by speakers from both Pittsburgh academia and industry. Dr. Robert Murphy, Dr. Carl Kingsford, Dr. Phillip Compeau, and Dr. Hosein Mohimani were among the speakers for these sessions.
Dr. Robert F. Murphy, Ray and Stephanie Lane Professor and Head of the Computational Biology Department, gave the Symposium’s keynote lecture. Dr. Murphy began by describing some of the previous paradigm shifts that biology has experienced, leading to the reductionist approach to research that dominated much of the twentieth century. He then discussed the shift from reductionism to “systems biology” and the dramatic shift from human data interpretation to computational modeling that is ongoing. He illustrated the need for this by showing machine performance outstripping human performance at identifying the patterns of cell organelles in microscope images, a task that twenty years ago was almost exclusively done visually. He went on to discuss how the demise of reductionism leads to the conclusion that the number of possible experiments that might be needed to understand even basic cell functions and behavior is beyond our ability to perform, and how a type of artificial intelligence method called “active machine learning” may hold the key to overcoming this hurdle. Dr. Murphy closed by discussing the growing number of jobs for people trained in computational biology at all levels, especially to realize the dream of preventative and therapeutic treatments customized to each individual.
Dr. Kingsford gave a presentation on the algorithms needed to piece together the sequence of a new genome from the short fragments that current DNA sequencers emit. He also discussed related problems where one seeks the sequences of expressed genes in order to find which variants of which genes are active in different conditions. His talk gave several examples of the interplay between ideas in computer science and graph theory that are needed to solve these biological problems.
Dr. Compeau also spoke on assembling genomes from tiny fragments of DNA, connecting the problem to that of finding a walk through the 18th Century city of Konigsberg that would cross every bridge exactly once. He also spoke on the frontier of computational biology, including the prospect of gene editing to revolutionize medicine, and how metagenomics will help us understand microscopic environments from the mutational landscapes of cancer tumors to the bacteria in our intestines. Finally, he touched on the power of machine learning to comprehend 20 petabytes worth of medical images in the US alone by 2020.