Parallel Algorithms for Inferring Gene Regulatory Networks: A Review

Area of Science:

• Systems Biology
• Bioinformatics
• Computational Biology

Background:

• Inferring large-scale gene regulatory networks from gene expression data presents significant computational challenges.
• Sequential algorithms are often infeasible for timely analysis of complex, high-dimensional biological datasets.

Purpose of the Study:

• To provide a comprehensive overview of recent parallel algorithms for constructing gene regulatory networks.
• To discuss the challenges associated with large-scale gene expression datasets in systems biology.

Main Methods:

• Review of fundamental concepts in gene regulatory network inference.
• Detailed examination of parallel computing frameworks: CUDA, OpenMP, MPI, and Hadoop.
• Analysis of various parallel algorithms designed for network construction.

Main Results:

• Identification of key parallel frameworks and libraries applicable to gene regulatory network inference.
• Overview of current parallel algorithmic approaches for building gene regulatory networks.
• Discussion of the feasibility and advantages of parallel computing for systems biology problems.

Conclusions:

• Massively parallel computing offers a viable solution for the timely inference of large-scale gene regulatory networks.
• The reviewed parallel frameworks and algorithms provide essential tools for systems biology research.
• Guidelines for parallel reverse engineering are presented to aid future research.