Prediction of seed gene function in progressive diabetic neuropathy by a network-based inference method

Area of Science:

• Genomics
• Bioinformatics
• Systems Biology

Background:

• Guilt by Association (GBA) algorithms are established for predicting gene functions.
• Network-based approaches enhance the reliability of identifying disease-specific molecular signatures.
• Progressive diabetic neuropathy (PDN) requires deeper understanding of its underlying molecular mechanisms.

Purpose of the Study:

• To develop and apply a novel network-based GBA method for identifying seed gene functions in PDN.
• To integrate the GBA algorithm with gene co-expression networks for improved prediction accuracy.
• To elucidate potential molecular pathways contributing to PDN progression.

Main Methods:

• Gene lists and sets were prepared, including 79 differentially expressed genes (DEGs) and 40 Gene Ontology (GO) terms.
• A co-expression matrix (CEM) was constructed using the Spearman correlation coefficient (SCC) method.
• Gene functions were predicted using the GBA algorithm, with seed functions selected based on the area under the receiver operating characteristics curve (AUC).

Main Results:

• The network-based GBA approach achieved good classification performance for 27.5% of gene sets (AUC >0.5).
• Three significant gene sets, with AUC >0.6, were identified as seed gene functions for PDN.
• These identified functions include binding, molecular function, and regulation of the metabolic process.

Conclusions:

• The study successfully predicted three seed gene functions critical for PDN progression using a network-based GBA algorithm.
• These findings offer valuable insights into the pathological and molecular mechanisms of PDN.
• The identified gene functions provide potential targets for future research and therapeutic strategies in PDN.