Identification and validation of Varroa-related hub genes in honey bees using combination of meta-analysis, weighted gene co-expression network analysis, and machine learning models

Varroa mites are a major threat to honey bee health, leading to heavy losses in colonies. The exact reason and mechanistic explanation of honey bee colony collapse be likely possible by the advent of transcriptomics data and investigation of host gene expression changes. In order to identify the mechanisms involved in the varroa disease process, different methods, including meta-analysis, weighted gene co-expression network analysis (WGCNA), functional enrichment analysis, gene ontology (GO), Kyoto Encyclopedia of Genes and Genomes (KEGG) pathway, attribute weighting and supervised machine learning model have been employed. Meta-analysis of three individual study has led to identify 454 (up=173, down=281) differentially expression genes (DEGs). WGCNA identified three significant functional modules associated with varroa disease. Additionally, GB51391, GB52250, GB44983, and GB40365 genes were identified as hub genes in the significant module networks, detected as most informative genes by different attribute weighting algorithms, and also identified as root genes in decision tree models. Systems biology analysis suggested that these DEGs play a key role in varroa disease by providing energy and improving immune system performance. This research provides new insights into the complex regulatory network of varroa disease processes and suggests several candidate genes that may be useful for future honey bee breeding programs. Furthermore, this study supports the idea that combining meta-analysis, WGCNA, and machine learning can achieve a higher resolution analysis and better predict the most important functional genes. This approach may provide a more robust bio-signature for disease traits and offer suitable biomarker candidates for future studies.