Explore a comprehensive lecture on applying data-driven model discovery methods in machine learning biology. Delve into topics such as chimeric antigen receptor T-cell immunotherapy, glioblastoma, and mathematical oncology. Learn about the intersection of mathematical modeling and machine learning, including concepts like Sparse Identification of Nonlinear Dynamics and Kepler's Planetary Motion. Examine case studies on modeling immortalized cell lines, the effects of dexamethasone, and CAR T-cell functional responses. Gain insights into model identification, latent variable identification, and the challenges of higher-order terms in modeling. Discover how these approaches can be applied to various dimensions of biological data, ultimately enhancing our understanding of complex biological systems and their potential applications in cancer research and treatment.
A Machine Learning Approach to Biology: Applying Data-Driven Model Discovery Methods
Mathematical Oncology via YouTube
Overview
Syllabus
Intro
Motivation
Chimeric antigen receptor Tcell immunotherapy
Glioblastoma
City of Hope
Clinical trials
Mathematical oncology
Immunotherapy
Chicken and the Egg
Mathematical modeling is an art
How modeling ends up going
How does mechanistic modeling
How does modeling and machine learning come together
What if you didnt have to do that
Sparse Identification Nonlinear Dynamics
Keplers Planetary Motion
What comes first
First paper
Modeling immortalized cell lines
What happens with dexamethasone
Dynamic phase faces
Rule change
Best model
Dolly
Model Identification
Tokens Theorem
Shadow manifolds
Latent variable identification
Library
Growth Depth
Car T Cell Functional Responses
Car T Cell Binding Dynamics
Minimal Accuracy
Models
Dual fit
Higher order terms
Data in code
Two Dimensions
Summary
Final Thought
Conclusion
Taught by
Mathematical Oncology
