Using First Principles to Reverse-engineer Complex Biological Systems
ENG-SCI 195
Subject & Catalog Number
Course Information
Description
As more modeling efforts are performed by machines, young scientists may become unfamiliar with intuitive reasoning. Here, I propose a thought-provoking class where students with quantitative background learn to develop intuitive thinking by constructing reductive hypotheses to characterize and predict the behavior of complex biological systems. All models rely on calculations made by hand without the help of any computer. Most of the tools used in the class are derived from advanced statistical physics. Students contribute during class to ‘live’ calculations initially derived by the instructor (black/white board) and then extended and challenged by peer-to-peer discussions in smaller groups. Learning outcome for students consists in being capable to apply first principles and simple reductive hypotheses for modeling dynamical behavior of molecular machines, stochastic intracellular networks, and large neural networks. They will also be capable to describe some physical limitations of machine learning approaches using Boolean networks. More generally, the underlying goal of the class is to demonstrate the usefulness of critical thinking as an essential tool to assess the validity of results from ‘black box’ modelling (in particular those produced by new generative AI platforms).
Course Notes
ES 195 is also offered as MCB 195. Students may not take both for credit.
Recommended Prep
Math 21a. Students who are familiar with multivariable calculus but did not take Math21a should consult with the instructor.
Available for Harvard Cross Registration
NOTE: This course requires additional sections; you will be prompted to choose secondary components during the Add to Cart process
