View in EDS

Multi-agent AI enables evidence-based cell annotation in single-cell transcriptomics

Saved in:
Bibliographic Details
Title: Multi-agent AI enables evidence-based cell annotation in single-cell transcriptomics
Authors: Ahuja, Gautam, Antill, Alex, Su, Yi, Dall'Olio, Giovanni Marco, Basnayake, Sukhitha, Karlsson, Göran, Dhapola, Parashar
Contributors: Lund University, Faculty of Medicine, Department of Laboratory Medicine, Division of Molecular Hematology (DMH), Stem Cells and Leukemia, Lunds universitet, Medicinska fakulteten, Institutionen för laboratoriemedicin, Avdelningen för molekylär hematologi, Stem Cells and Leukemia, Originator, Lund University, Faculty of Medicine, Department of Laboratory Medicine, Division of Molecular Hematology (DMH), Lunds universitet, Medicinska fakulteten, Institutionen för laboratoriemedicin, Avdelningen för molekylär hematologi, Originator, Lund University, Profile areas and other strong research environments, Strategic research areas (SRA), StemTherapy: National Initiative on Stem Cells for Regenerative Therapy, Lunds universitet, Profilområden och andra starka forskningsmiljöer, Strategiska forskningsområden (SFO), StemTherapy: National Initiative on Stem Cells for Regenerative Therapy, Originator
Subject Terms: Medical and Health Sciences, Basic Medicine, Basic Cancer Research, Medicin och hälsovetenskap, Medicinska och farmaceutiska grundvetenskaper, Basal cancerforskning
Description: Cell type annotation remains a critical bottleneck, with current methods often inaccurate and requiring extensive manual validation, particularly in disease contexts. While large language models (LLMs) show promise, they can be unreliable due to hallucinations. We developed CyteType, a multi-agent framework that generates competing hypotheses grounded in full expression data and study context, validates against external databases, and iteratively self-evaluates. Comprehensive benchmarking demonstrates that CyteType substantially outperforms reference-based and LLM-based methods, with self-generated confidence scores reliably identifying trustworthy annotations. CyteType transforms cell type annotation from label assignment into evidence-grounded biological discovery.
Access URL: https://doi.org/10.1101/2025.11.06.686964
Database: SwePub
Description
Abstract:Cell type annotation remains a critical bottleneck, with current methods often inaccurate and requiring extensive manual validation, particularly in disease contexts. While large language models (LLMs) show promise, they can be unreliable due to hallucinations. We developed CyteType, a multi-agent framework that generates competing hypotheses grounded in full expression data and study context, validates against external databases, and iteratively self-evaluates. Comprehensive benchmarking demonstrates that CyteType substantially outperforms reference-based and LLM-based methods, with self-generated confidence scores reliably identifying trustworthy annotations. CyteType transforms cell type annotation from label assignment into evidence-grounded biological discovery.
DOI:10.1101/2025.11.06.686964