Computational Systems & Biology
We are standing at a major inflection point for data and biomedical science—the way we view and practice scientific research is changing profoundly. These changes are being driven by computational systems biology, an interdisciplinary and data-driven approach to biomedicine, which will increasingly transform biomedicine from disease-driven and reactive to health-driven and predictive, yet preventative. ECE’s systems biology researchers work with biologists, chemists, and clinical researchers to develop experiments and mathematical and computational models, along with underlying theory and software tools.
William T. Baumann
Yue (Joseph) Wang
Cancer treatments tend to work well for a period of time and then cease being ben-eficial once resistance develops. An ECE team is building mathematical models of breast cancer cells responding to different targeted therapies in an effort to figure out how to use combinations of drugs in a se-quence that prevents the development of resistance. The model will be used to help doctors optimize therapies and provide clinical insight.
In this project, we seek to identify what drives breast cancer growth and determine how to stop it. We are applying computational modeling methods, using principles from machine learning and ecology, to study how therapy-resistant and sensitive cells cooperate to alter the response of cancer cells to treatment. This knowledge will be used to identify new interventions, or optimizations of existing regimens, to improve outcomes for patients.
Astrocytes are the most numerous glial cells and are estimated to outnumber neurons in the brain. They are deeply involved in normal brain development and in brain disease. Current methods of analyzing astrocyte activity data is essentially manual. ECE researchers have developed a suite of computational tools that automatically quantify the functional status of astrocytes. To understand astrocyte-calcium signal data, we invented a new mathematical theory for joint alignment of multiple curves, called graphical time warping. Our approach transforms the joint alignment problem into a network flow problem that can be exactly and efficiently solved.
The disease process that causes heart attacks involves an abnormal collection of proteins in artery walls. An ECE team is working to identify these proteins and to discover why they accumulate and how to prevent them from doing so. We are applying multiple analysis methods to deconvolve complex proteomic signals to provide detailed descriptions of the arterial proteome and network re-wiring in atherosclerosis. We will use the knowledge gained to inform the design of an early disease detection assay and hope to help reduce the burden of atherosclerosis.