Machine Learning

Our DNA contains genes that provide instructions for making proteins essential for life. However, genes alone aren't enough – they need switches that control when and where they turn on or off. These switches, called regulatory elements, are scattered throughout our DNA, often far from the genes they control. About 650 000 potential regulatory elements have been identified in human DNA. However, we don't understand how they choose which genes to control or how they work. This is important because genetic changes in regulatory elements can lead to diseases, including cancer.

Develop and apply new algorithms and machine learning methods for analyzing proteomic data and integrating multi-omics data (including genomics, transcriptome, other MS data like metabolomics and lipidomics, and metagnomics).
 

I'm working on developing mathematical and computational tools to analyze high-dimensional genomics data and model development in living organisms.

As a member of the Canadians Saving Arrest Victims Everywhere (CanSAVE) Novel Biosensor project, Mahsa's research is focused on developing wearable technologies to detect out-of-hospital sudden cardiac arrest incidents. This work involves (1) designing wearable sensors to collect bio-signals associated with cardiac arrest conditions (e.g., electrocardiogram, breathing rate); and (2) using machine learning to identify appropriate combinations of collected bio-signals to detect a sudden cardiac arrest event.