Benchmarking imputation methods for network inference using a novel method of synthetic scRNA-seq data generation.

Area of Science:

• Computational Biology
• Systems Biology
• Genomics

Background:

• Single-cell RNA sequencing (scRNA-seq) offers deep insights into cellular heterogeneity.
• Gene expression correlations in scRNA-seq data can reveal gene regulatory networks.
• Technical biases, such as drop-out events, complicate scRNA-seq data interpretation.

Purpose of the Study:

• To introduce Biomodelling.jl, a novel tool for generating synthetic scRNA-seq data.
• To provide a benchmark for evaluating gene regulatory network inference algorithms.
• To investigate the impact of zero-imputation methods on network inference.

Main Methods:

• Utilized multiscale modeling of stochastic gene regulatory networks in dynamic cellular environments.
• Developed Biomodelling.jl for the creation of synthetic scRNA-seq transcription data.
• Generated data with known ground truth network topology for benchmarking.

Main Results:

• Biomodelling.jl successfully produces realistic scRNA-seq data.
• The tool enables systematic benchmarking of network inference approaches.
• Investigated the influence of various imputation techniques on algorithm performance.

Conclusions:

• Biomodelling.jl is a valuable resource for developing and validating scRNA-seq based network inference methods.
• Facilitates rigorous assessment of data processing strategies in single-cell analysis.
• Aids in advancing the accuracy and reliability of gene regulatory network reconstruction.