Algorithm Helps Unlock Secrets of Cell Biology

A novel computer algorithm gave a big boost to cell biology research recently when UT Dallas researchers teamed with the Computational Biology Program at Memorial Sloan-Kettering Cancer Center in New York.

Researchers have set the algorithm, called MEDUSA, loose on a compelling biological question—namely, what genes control cells when the air gets switched off?

MEDUSA has helped process existing biological data to make new predictions in the study of disease and oxygen sensing.

MEDUSA, short for motif element discrimination using sequence agglomeration, uses a “machine learning” approach – extracting rules or patterns from massive data sets – to tease patterns out of existing data.  Aging, certain diseases,  and environmental toxins can lower oxygen levels within cells (called hypoxia), and the results on the body can be disastrous.

Dr. Li Zhang, the Cecil H. and Ida Green Distinguished Chair in Systems Biology and head of the UT Dallas Department of Molecular and Cell Biology and her co-author, Dr. Christina Leslie, assistant member of the Sloan-Kettering Institute at Memorial Sloan-Kettering Cancer Center, have published an article about their findings in the recent issue of PLOS Computational Biology.

Zhang supplied the microbiological data and expertise, while Leslie brought MEDUSA to bear on gene data from microarray experiments—a genetic “snapshot” of the current state of the cell.

“Before scientists in our field began to team like this – microbiologist with computational biologist – it was like seeing just a small piece of the sky,” Zhang said.  “Now, thanks to our partnership with Memorial-Sloan Kettering Cancer Center, we can see the whole sky.”

Zhang uses yeast cells – the same species that make beer – as a model to study the genes that control how cells use oxygen, or how cells respond in low oxygen environments, like mountainous regions.

Leslie, whose team developed MEDUSA, said teamwork between her lab and Zhang’s was critical to figuring out whether an algorithm can explain what happens in cells.

“What was exciting about the collaboration was that after we learned the computational model, Li’s group was able to go back and experimentally confirm a set of our predictions,” Leslie said.  “A lot of the computational work was done through close interaction with Li, to determine what kinds of analysis and predictions might be most meaningful from a biological viewpoint and most appropriate for experimental follow-up.”

The published paper, A Predictive Model of the Oxygen and Heme Regulatory Network in Yeast, was published in the official journal of the International Society for Computational Biology’s November issue.

Cancer research, studies about brain health and stroke recovery, and a wide array of other research avenues could ultimately benefit from the sophisticated one-two punch that comes with pairing microbiology with computational biology.