What question did this study set out to answer?

This research aims to create a standardized and accurate framework for cell type annotation in single-cell RNA sequencing analysis.

April 19, 2026

Abstract LB169: scCAP: An ontology-aware large language model framework for hierarchical and standardized single-cell type annotation

Key Points

This research aims to create a standardized and accurate framework for cell type annotation in single-cell RNA sequencing analysis.
Developed scCAP using an ontology-aware large language model architecture
Integrated multiple annotation tools through a meta-learning framework
Employed HybridAttentionNet with knowledge and confidence streams
Utilized LLM for structured descriptions of label predictions
Evaluated against existing annotation tools using cell atlas data
scCAP achieved 0.898 average semantic similarity at Level 3 annotation
At 80% similarity threshold, achieved 87.7% accuracy versus 73.3% for CellTypist
Demonstrated consistent hierarchical performance across levels with higher median similarity
Tool dropout regularization ensured robust performance even with missing tools

Abstract

Abstract Background: Accurate cell type annotation remains challenging in single-cell RNA sequencing analysis. Existing tools (CellTypist, SCimilarity, SingleR, Geneformer) exhibit distinct biases and lack standardized nomenclature, hindering cross-study comparability. Furthermore, current methods fail to provide hierarchical annotations reflecting cell type ontologies. Methods: We developed scCAP, a meta-learning framework integrating multiple annotation tools using an ontology-aware large language model (LLM) architecture. scCAP employs HybridAttentionNet combining: (1) a knowledge stream using self-attention on semantic embeddings, and (2) a confidence stream learning tool-specific reliability patterns. A key innovation is our LLM-based label interpretation: for each prediction, the LLM generates structured descriptions including full names, marker genes, and pathways, enabling automatic abbreviation expansion and cluster annotation interpretation. These descriptions are embedded for standardized semantic comparison across heterogeneous nomenclatures. The LLM also classifies labels into hierarchical ontology levels (L1: lineages; L2: cell classes; L3: subtypes), validated against Cell Ontology. Tool dropout regularization during training ensures robust performance even when individual tools are unavailable. Additionally, iterative pseudo-labeling progressively refines predictions by incorporating high-confidence annotations as training signals. The modular architecture allows seamless tool addition/replacement without retraining. Results: We evaluated scCAP on Cancer Cell Census Atlas and Human Lung Cell Atlas data against four annotation tools. At Level 3, scCAP achieved 0. 898 average semantic similarity, outperforming SingleR (0. 865), CellTypist (0. 861), SCimilarity (0. 857), and Geneformer (0. 805). At ≥80% similarity threshold, scCAP achieved 87. 7% accuracy versus 73. 3% for CellTypist. scCAP demonstrated consistent hierarchical performance (L1: 0. 858; L2: 0. 874; L3: 0. 898) with higher median similarity and tighter variance than individual tools. Conclusions: scCAP leverages LLMs for ontology-aware label standardization and hierarchical classification. Its plug-and-play architecture enables incorporating emerging tools while maintaining consistent outputs, facilitating meta-analyses of tumor microenvironment heterogeneity. Citation Format: Dongkwan Shin, Seok-Won Jang, Jonghyun Lee, Juyeon Cho. scCAP: An ontology-aware large language model framework for hierarchical and standardized single-cell type annotation abstract. In: Proceedings of the American Association for Cancer Research Annual Meeting 2026; Part 2 (Late-Breaking, Clinical Trial, and Invited Abstracts) ; 2026 Apr 17-22; San Diego, CA. Philadelphia (PA): AACR; Cancer Res 2026;86 (8Suppl): Abstract nr LB169.

AIに質問

Bookmark