Single-cell technologies are transforming biology by allowing researchers to analyze individual cells rather than relying on averaged data. However, these approaches face challenges like high costs and balancing resolution, throughput, and spatial information. Computational tools are helping to overcome these barriers, enabling groundbreaking discoveries like those showcased by the Human Cell Atlas. Among these, tools developed by Dr. Jun Ding and his lab at McGill and the RI-MUHC — scSemiProfiler, UNAGI, and CellAgentChat — are pivotal advancements in the field.
Dr. Jun Ding’s work was recognized in the Nature Technology Features, highlighting its transformative potential for single-cell research and therapeutic discovery. These tools democratize single-cell omics, integrate AI to simulate cellular processes, and enhance the scalability of studies, addressing limitations in cost, throughput, and clinical translation.
Nature credits Dr. Jun Ding’s contributions as crucial for advancing the the Human Cell Atlas’s goals and paving the way for breakthroughs in precision medicine and drug development. As single-cell technologies continue to grow, tools like these are expected to play an even larger role in biological discovery.
scSemiProfiler:
This tool addresses the high cost of single-cell RNA sequencing by leveraging bulk RNA sequencing data with generative AI to reconstruct high-resolution single-cell profiles. Ding’s team demonstrated that scSemiProfiler could reduce sequencing costs by approximately 80%, making single-cell studies more accessible. For example, they used it to analyze immune cells from COVID-19 patients, generating accurate profiles with minimal single-cell data.
scSemiProfiler: Advancing large-scale single-cell studies through semi-profiling with deep generative models and active learning. Wang J, Fonseca GJ, Ding J. Nat Commun. 2024 Jul 16;15(1):5989.
UNAGI:
Designed for in silico drug discovery, UNAGI models disease progression at the cellular level. Ding’s lab used this tool to create a “sandbox” for idiopathic pulmonary fibrosis (IPF), simulating how cells evolve during disease. The model identified existing drugs like nintedanib while flagging new therapeutic candidates. This tool shows promise in revolutionizing drug discovery by simulating the effects of treatments before physical experiments.
Unagi: Deep Generative Model for Deciphering Cellular Dynamics and In-Silico Drug Discovery in Complex Diseases. Zheng Y, Schupp J, Adams T, Clair G, Justet A, Ahangari F, Yan X, Hansen P, Carlon M, Cortesi E, Vermant M, Vos R, D Sadeleer L, Rosas I, Pineda R, Sembrat J, Konigshoff M, Mcdonough J, Vanaudenaerde B, Wuyts W, Kaminski N, Ding J. Preprint at Research Square https://doi.org/10.21203/rs.3.rs-3676579/v1 (2023).
CellAgentChat:
Unlike conventional approaches, this tool models individual cells as autonomous agents that interact dynamically within their environments. By simulating these interactions, CellAgentChat enables researchers to predict cell-cell signaling pathways and test drug effects in silico. For example, using breast cancer data, the model identified key interactions and drug targets like the epidermal growth factor receptor.
Harnessing Agent-Based Modeling in CellAgentChat to Unravel Cell-Cell Interactions from Single-Cell Data. Raghavan V, Li Y, Ding J. Preprint at bioRxiv https://doi.org/10.1101/2023.08.23.554489 (2024).