ELeFHAnt

ELeFHAnt annotates and harmonizes single-cell RNA sequencing (scRNA-seq) data by leveraging ensemble learning to integrate multiple reference Seurat objects into unified cellular atlases.

Key Features:

• CelltypeAnnotation: Annotates cells in a query Seurat object using a reference Seurat object with pre-annotated cell types via support vector machine and random forest classifiers.
• LabelHarmonization: Harmonizes cell type labels across multiple reference Seurat objects to generate a unified cellular atlas for integrated analysis.
• DeduceRelationship: Compares and deduces relationships between cell types across two scRNA-seq datasets to reveal similarities and differences in cellular composition.
• Ensemble learning: Combines support vector machine (SVM) and random forest algorithms to improve annotation accuracy and robustness across supervised and unsupervised contexts.

Scientific Applications:

• Cell type annotation: Assigns cell type labels to cells in scRNA-seq datasets using annotated reference Seurat objects.
• Reference integration: Integrates multiple scRNA-seq reference atlases into a cohesive cellular atlas for comparative studies.
• Cross-dataset comparison and meta-analysis: Enables comparison of cellular compositions between independent scRNA-seq studies to support meta-analysis and interpretation.

Methodology:

ELeFHAnt employs ensemble learning by combining support vector machine (SVM) and random forest algorithms for cell type annotation, providing robust performance in supervised and unsupervised contexts.