EnTSSR: A Weighted Ensemble Learning Method to Impute Single-Cell RNA Sequencing Data.

Area of Science:

• Genomics
• Bioinformatics
• Computational Biology

Background:

• Single-cell RNA sequencing (scRNA-seq) offers deep insights into cellular states and disease mechanisms.
• scRNA-seq data is prone to technical dropout events, leading to false zero counts and impacting downstream analysis.
• Existing computational methods for imputing dropout events lack a strong consensus on optimal performance.

Purpose of the Study:

• To introduce EnTSSR, a novel weighted ensemble learning method for imputing dropout events in scRNA-seq data.
• To leverage consensus similarities between genes and cells using a multi-view sparse self-representation framework.
• To effectively integrate information from various imputation methods via a weighted ensemble strategy.

Main Methods:

• Developed EnTSSR, a weighted ensemble learning approach.
• Utilized a multi-view two-side sparse self-representation framework.
• Evaluated performance through down-sampling, clustering, differential expression, and cell trajectory inference.

Main Results:

• EnTSSR effectively recovers true expression patterns in scRNA-seq data.
• The weighted ensemble strategy successfully leverages information from multiple imputation methods.
• Experimental results validate the model's capability in handling dropout events.

Conclusions:

• EnTSSR provides a robust solution for scRNA-seq data imputation.
• The method enhances the reliability of downstream analyses in complex disease research.
• EnTSSR represents a significant advancement in scRNA-seq data processing.