Therapeutic design through the lens of data-driven biomolecular systems engineering

There is serious concern that the continued increase in drug development costs observed in recent decades is unsustainable and will lead to further increases in already exorbitant drug prices, highlighting the need for a new paradigm for therapeutic discovery. To meet these challenges, our group develops tools for therapeutic design to accelerate discovery and enable personalized medicine. Our work integrates molecular and systems modeling, machine learning, and global optimization to develop multi-scale models of biomolecular systems, and then leverages these models for the design of novel therapeutics. In this talk, I will highlight our recent work that has focused on developing methods for training and analyzing explainable or interpretable machine learning models aimed at predicting therapeutically relevant peptide properties and function based on sequence alone. Though recent advances in deep learning have transformed the field of bioinformatics, there is still great need for explainable data-driven models that can connect biomolecular descriptors to biomolecular function, and ultimately to health and disease. For instance, explainable machine learning models can be used for understanding the physicochemical driving forces of biomolecular function, to identify novel biomarkers, and serve as guides for the design of novel therapeutics. We have been developing an approach that uses the Fourier transform to measure the oscillations of physicochemical properties of amino acids along peptide sequences. These Fourier-based encodings have major advantages since they encode amino acid order and can be used to encode peptides with varied length without the need of sequence alignments. One approach for incorporating interpretability is feature selection, which allows for the identification of an optimal subset of features that can improve model performance. We can also use feature selection to help elucidate the physicochemical properties that are most important in predicting a desired property or phenomena. Our approach allows us to profile and quantify functional similarities between classes of biomolecules and connect biomolecular differences/similarities in function to physicochemical properties.

Biography

Dr. Chris A. Kieslich earned his B.S. in Biomedical Engineering from Saint Louis University (St. Louis, MO) in 2007 and his Ph. D. in Bioengineering from University of California, Riverside (Riverside, CA) in 2012. Dr. Kieslich did postdoctoral research in Chemical Engineering with Christodoulos Floudas at Princeton University (Princeton, NJ) from 2012-2014 and at Texas A&M University (College Station, TX) from 2014-2016 after Dr. Floudas relocated to become the Director of the Texas A&M Energy Institute. He was Research Faculty in the Biomedical Engineering at Georgia Institute of Technology (Atlanta, GA) from 2016-2019 before joining the faculty in the Department of Chemical Engineering at Auburn University as an Assistant Professor in 2019 and has since rejoined the Walter H. Coulter Department of Biomedical Engineering at Georgia Institute of Technology as an Assistant Professor in 2025. His research is focused on the development and application of computational tools for biomolecular systems modeling with particular interest in therapeutic design. He is a recipient of the Maximizing Investigators’ Research Award (MIRA) for Early-Stage Investigator from the National Institutes of Health and was named a Ginn Faculty Achievement Fellow by the Samuel Ginn College of Engineering in 2022.