Learning and Designing Interventions
Name
zhang-viczhang-phd-eecs-2026-thesis.pdf
Size
49.02 MB
Format
Adobe PDF
Checksum (MD5)
aa148199650ba22693bf2ebcdbd6e13e
Author(s)
Zhang, Jiaqi
Advisor(s)
Uhler, Caroline
Date Issued
May 2026
Publisher
Massachusetts Institute of Technology
Abstract
Complex causal mechanisms among genes govern cellular functions in health and disease. Understanding these mechanisms can accelerate therapeutic discovery but remains challenging due to the large number of genes and their intricate dependencies. Recent advances in experimental technologies are making this problem increasingly tractable: it is now possible to systematically intervene on individual genes or gene combinations in single cells and measure their downstream effects, enabling empirical identification and validation of causal relationships. However, interventional data are high-dimensional, making interpretation challenging, and costly to collect. In this thesis, I present our work tackling these challenges in three parts. In the first part, we develop efficient algorithms for causal discovery when the causal variables are observed, addressing the bottleneck of using too many conditional independence tests in existing algorithms. In the second part, we introduce theories and methods for learning causal representations when the causal variables are not directly observed. We also demonstrate how to predict the effects of novel interventions using these representations. Finally, in the third part, we propose frameworks for optimizing intervention designs and showcase that these frameworks can be useful for and applied to biomedical problems in disease-relevant contexts.
MIT Department
Massachusetts Institute of Technology. Department of Electrical Engineering and Computer Science
Terms of Use
In Copyright - Educational Use Permitted
Copyright retained by author(s)
Persistent DSpace Link
