Research

I'm interested in building interpretable machine learning methods for biology, using ideas from probability, information theory, and statistical physics to make models more biologically meaningful.

Currently, in the Wagner lab, I work on spatial metabolism, studying how a cell's metabolism is shaped by its surrounding tissue environment. I've also worked on MeRN, a framework for inferring interpretable metabolic activity from single-cell RNA-seq.

Previously, I've worked on quantum machine learning for small organic molecule discovery and computational microscopy in the Tian Lab.

Selected Publications

See also Google Scholar and ORCID.

• Extracting interpretable single-cell metabolic states with graph-guided representation learning

  Daniel P. Lewinsohn, Nicolas Dias, Adelina Chau, Yuko Koike, Zachary D. Smith, Nilah M. Ioannidis, Allon Wagner

  bioRxiv preprint, 2026 [bioRxiv][code]

• Folate deficiency disrupts key metabolic transitions within the developing neural ectoderm

  Nicolas Dias, Daniel P. Lewinsohn, William N. Colgan, Minming Wang, Yusuke Kijima, JoAnne Villagrana, Tien-Chi Jason Hou, Gokul Gowri, Adelina Chau, Tuğçe Aktaş, Kaelyn Sumigray, Jonathan S. Weissman, Luke W. Koblan, Allon Wagner, Zachary D. Smith

  bioRxiv preprint, 2026 [bioRxiv]

• Multi-channel FourierNet for large-scale shift variant reconstruction

  Qianwan Yang, Ruipeng Guo, Guorong Hu, Adelina Chau, Jamin Xie, Lei Tian

  SPIE Computational Optical Imaging and Artificial Intelligence in Biomedical Sciences, 2024 [doi]