Progress and challenges in the computational prediction of gene function using networks: 2012-2013 update

Area of Science:

• Bioinformatics
• Computational Biology
• Systems Biology

Background:

• Gene network-based guilt-by-association (GBA) methods are widely used for inferring gene function.
• Previous work highlighted that gene multifunctionality acts as a significant confound in GBA analyses.
• The reliance on gold-standard annotations like Gene Ontology may limit the discovery of novel functional insights.

Purpose of the Study:

• To review recent advancements in guilt-by-association research.
• To discuss ongoing challenges and new directions in the field.
• To re-evaluate the impact of gene multifunctionality on GBA.

Main Methods:

• Review of recent literature on guilt-by-association (GBA) and gene function prediction.
• Analysis of studies addressing confounding factors in computational biology.
• Discussion of the role of gene multifunctionality and annotation databases.

Main Results:

• Gene multifunctionality continues to significantly impact the reliability of GBA signals.
• Evaluating and applying computational GBA remains more challenging than previously assumed.
• Progress has been made in understanding these challenges, with new research avenues emerging.

Conclusions:

• Efforts in computational biology should prioritize identifying data features that provide novel functional insights.
• Rethinking the reliance on traditional annotation databases is crucial for advancing GBA.
• Continued research is needed to overcome the inherent challenges in gene network analysis.