New foundation model Tabula uses federated learning for privacy-preserving single-cell genomics

Researchers have developed Tabula, a novel foundation model for single-cell genomics that addresses privacy concerns by utilizing federated learning. This model explicitly handles the tabular structure inherent in single-cell data, unlike previous approaches. Tabula, deployed via the Chiron platform, enables collaborative training across institutions without raw data sharing. The model demonstrates strong performance on various biological benchmarks, uncovering regulatory logic in systems like hematopoiesis and neurogenesis, and identifies potential rejuvenation factors for human fibroblasts. AI

IMPACT Enhances privacy in biological data analysis and advances foundation models for tabular biological data.

RANK_REASON The cluster describes a new scientific paper detailing a novel foundation model for a specific scientific domain. [lever_c_demoted from research: ic=1 ai=1.0]

Read on arXiv cs.LG →

• cardiogenesis
• Chiron
• federated learning
• human fibroblast
• neurogenesis
• pancreatic endogenesis
• single-cell RNA-seq
• Tabula

• paper
• infra
• safety

AI-generated summary · Google Gemini · from 1 sources. How we write summaries →