Related Experiment Video
Updated: May 27, 2025
10:36
Rare Event Detection Using Error-corrected DNA and RNA Sequencing
Published on: August 3, 2018
12.0K
PbImpute: Precise Zero Discrimination and Balanced Imputation in Single-Cell RNA Sequencing Data
Yi Zhang1,2, Yin Wang1,2, Xinyuan Liu1,2
1School of Computer Science and Engineering, Guilin University of Technology, 12 Jiangan Road, Qixing District, Guilin 541004, China.
Journal of Chemical Information and Modeling
|February 17, 2025
Summary
PbImpute accurately imputes single-cell RNA sequencing (scRNA-seq) data by balancing dropout recovery and biological zero preservation. This method enhances cellular heterogeneity analysis and disease research by improving data fidelity.
Area of Science:
- Genomics and Bioinformatics
- Computational Biology
Background:
- Single-cell RNA sequencing (scRNA-seq) reveals cellular heterogeneity but suffers from technical "dropout zeros".
- Existing imputation methods often cause under- or over-imputation, distorting scRNA-seq data interpretation.
Purpose of the Study:
- To develop a precisely balanced imputation (PbImpute) method for scRNA-seq data.
- To achieve optimal equilibrium between recovering missing data and preserving true biological zeros.
Main Methods:
- PbImpute integrates zero-inflated negative binomial (ZINB) modeling with static and dynamic repair algorithms.
- It employs multistage imputation including parameter optimization, static repair, secondary dropout identification, graph-embedding neural networks, and dynamic repair.
Main Results:
- PbImpute demonstrated superior performance in discriminating technical dropouts from biological zeros (F1 Score = 0.88).
- The method significantly improved gene-gene and cell-cell correlations, differential expression analysis, and clustering visualization.
- Ablation studies confirmed the effectiveness of both imputation and repair modules.
Conclusions:
- PbImpute offers a balanced approach to scRNA-seq data imputation, reducing signal distortion and preserving biological signals.
- This method enhances the accuracy of cell subpopulation identification and differential gene expression analysis, advancing cellular heterogeneity studies and disease research.
