Project Description:

Machine learning (ML) has permeated into all aspects of engineering and medicine. However, use of ML as black box tools can result in unreliable models and lacks interpretability. This projects focuses on ML multiscale modeling of complex biological systems during development. Fundamental understanding and new ML algorithm design is needed to combine data from different scales (individual cell to the entire embryo) and different species (zebrafish and drosophila) to produce reliable models that can explain the data, take into account known biological and physical processes, offer interpretable and testable hypotheses, can deal with noise and uncertainty, and are computationally efficient. Development of these new tools has a potential for revolutionizing biophysics research.

Start Date:

07/01/2023

Postdoc Qualifications:

Machine learning, Computational physics

Co-Advisors:

Adrian Buganza Tepole, Mechanical Engineering
David Umulis, Biomedical Engineering

Bibliography:

Alber M, Buganza Tepole A, Cannon WR, De S, Dura-Bernal S, Garikipati K, Karniadakis G, Lytton WW, Perdikaris P, Petzold L, Kuhl E. Integrating machine learning and multiscale modeling—perspectives, challenges, and opportunities in the biological, biomedical, and behavioral sciences. NPJ digital medicine. 2019 Nov 25;2(1):1-1.

Tac V, Costabal FS, Tepole AB. Data-driven tissue mechanics with polyconvex neural ordinary differential equations. Computer Methods in Applied Mechanics and Engineering. 2022 Aug 1;398:115248.

Burzawa L, Li L, Wang X, Buganza-Tepole A, Umulis DM. Acceleration of PDE-based biological simulation through the development of neural network metamodels. Current pathobiology reports. 2020 Dec;8(4):121-31.