Determination of core pathways for oral squamous cell carcinoma via the method of attract.

Objective: We expected to demonstrate a practical framework for oral squamous cell carcinoma (OSCC) candidate biomarker analysis at the pathway level based on the attract method, so as to give great insights to reveal the pathological mechanism underlying this disease at its early stage.

Methods: First, gene expression profile of OSCC was recruited and preprocessed. Then, Kyoto encyclopedia of genes and genomes (KEGG) pathway enrichment analysis was conducted. Next, attract method, an approach that begins its analysis from the "foundation knowledge sets" to discriminate the cell-phenotypes by those well-annotated gene-sets, then expands the syn-expression groups via decomposing each significant pathway into correlated subsets and extends the analysis to the entire expression was applied to identify core pathways. Finally, gene ontology (GO) functional enrichment analysis was performed on each of the correlated set groups to discover any potentially shared biological themes.

Results: A total of 226 pathways were obtained. Then, 39 core KEGG pathways was identified via attract. After removing the uninformative genes, a total of 1, 2, and 3 clusters were separately identified for the three discriminative pathways extracellular matrix (ECM)-receptor interaction, neuroactive ligand-receptor interaction, and cell adhesion molecules (CAMs) pathway based on the correlation coefficient < 0.85. GO functional enrichment analysis for the correlated partners groups indicated that there were 40, 11, 78 significant GO terms for ECM-receptor interaction, neuroactive ligand-receptor interaction, and CAMs pathway, respectively.

Conclusions: We predict that pathways such as ECM-receptor interaction, neuroactive ligand-receptor interaction, and CAMs may play significant roles in OSCC and targeting these pathways may provide an effective avenue to combat the complicated illness.