Singletrome enhances detection of long noncoding RNAs in single cell transcriptomes

Area of Science:

• Genomics
• Bioinformatics
• Molecular Biology

Background:

• Single-cell RNA sequencing (scRNA-seq) primarily focuses on protein-coding genes, overlooking the diversity of long noncoding RNAs (lncRNAs).
• Existing reference annotations for scRNA-seq underrepresent lncRNAs and merging them can lead to inaccurate read counts due to overlapping gene structures.

Purpose of the Study:

• To introduce Singletrome, a tool for generating comprehensive gene annotations that include lncRNAs.
• To enable accurate mapping and analysis of scRNA-seq data using these enhanced annotations.
• To facilitate the discovery and utilization of lncRNAs in cell type identification and disease prediction.

Main Methods:

• Development of Singletrome, a Singularity image integrating protein-coding and lncRNA gene transfer format (GTF) annotations.
• Generation of enhanced annotations accounting for sense and antisense gene overlap.
• Mapping scRNA-seq data and producing downstream analysis files.

Main Results:

• Detection of thousands of novel lncRNAs beyond current GENCODE annotations.
• Successful clustering of cell types based exclusively on lncRNA expression profiles.
• Demonstration of machine learning models predicting cell type and disease using lncRNAs.

Conclusions:

• Singletrome provides comprehensive lncRNA annotations for scRNA-seq analysis.
• lncRNAs are valuable biomarkers for cell type and disease state identification.
• This work facilitates the creation of a human lncRNA atlas in health and disease.