Facilitate integrated analysis of single cell multiomic data by binarizing gene expression values.

Area of Science:

• Genomics
• Computational Biology
• Molecular Biology

Background:

• Cell identity is defined by dynamic transcriptome and epigenome profiles, leading to diverse cell states.
• Standard single-cell RNA sequencing (scRNA-seq) analysis uses gene expression read counts.
• Binarization of gene expression (classifying genes as "on" or "off") offers an alternative for cell clustering.

Purpose of the Study:

• To demonstrate the effectiveness of combining binarized scRNA-seq and single-cell ATAC sequencing (scATAC-seq) data for integrated clustering.
• To present a method that bypasses the need to convert scATAC-seq data into gene activity scores.
• To enable direct assessment of each data modality's contribution to cell type resolution.

Main Methods:

• Applied term-frequency-inverse document frequency (TF-IDF) and singular value decomposition (SVD/LSI) to combined data.
• Utilized binarized scRNA-seq data and standard scATAC-seq data.
• Performed vertical integrated clustering on paired multiomic data.

Main Results:

• Direct concatenation of binarized scRNA-seq and scATAC-seq data is sufficient for effective vertical clustering.
• The proposed method avoids scATAC-seq to gene activity score conversion.
• Facilitates direct investigation into the contribution of each modality for cell type identification.

Conclusions:

• Combined binarized scRNA-seq and scATAC-seq data offer a powerful approach for multiomic cell type analysis.
• This method simplifies integrated clustering and enhances cell state resolution.
• The approach provides insights into the distinct roles of transcriptomic and epigenomic data in defining cell identity.