Publications

Linker-Length Landscape Mapping Enables Coupling of Diverse Synthetic Chemically Induced Dimerization Systems to Molecular Readouts

Published in bioRxiv (Preprint), 2026

We map how linker length couples diverse synthetic chemically induced dimerization (CID) systems to fluorescent sensor readouts, yielding sensors with dynamic ranges up to 1270% in mammalian cells.

Pan Y, Kang S, **Nakajima An D**, Yu Y, DiMaio F, Gu L. Linker-Length Landscape Mapping Enables Coupling of Diverse Synthetic Chemically Induced Dimerization Systems to Molecular Readouts. bioRxiv; 2026. doi: https://doi.org/10.64898/2026.07.01.735888 https://www.biorxiv.org/content/10.64898/2026.07.01.735888v1

AF2Complex predicts direct physical interactions in multimeric proteins with deep learning

Published in Nature Communications, 2021

This paper is about our work extending AlphaFold2 to multimeric proteins.

Gao, M., **Nakajima An, D.**, Parks, J.M. et al. AF2Complex predicts direct physical interactions in multimeric proteins with deep learning. Nat Commun 13, 1744 (2022). https://doi.org/10.1038/s41467-022-29394-2 https://rdcu.be/cKnut

Deep Learning System for Labeling Neurology Text

Published in 146th Annual Meeting of the American Neurological Association, 2020

In this poster presentation, we present the application of an interactively trained labeling-rule-based DL system to classify clinical and scientific text in neurology fields.

**An, Davi Nakajima**, Kartchner, D., Zhang, C., & Mitchell, C. S. (2021, October). Deep Learning System for Labeling Neurology Text for Predictive Medicine. In ANNALS OF NEUROLOGY (Vol. 90, pp. S159-S160). 111 RIVER ST, HOBOKEN 07030-5774, NJ USA: WILEY.

ReGAL: Rule-Generative Active Learning for Model-in-the-Loop Weak Supervision

Published in NeurIPS 2020 Workshop on Human And Model in the Loop Evaluation and Training Strategies, 2020

In this workshop paper, we present our interactive weak supervision model for text classification

David Kartchner, Wendi Ren, **Davi Nakajima An**, Chao Zhang, and Cassie S. Mitchell. ReGAL: Rule-Generative Active Learning for Model-in-the-Loop Weak Supervision. Retrieved from https://par.nsf.gov/biblio/10286627. Advances in neural information processing systems .