Authentication of differential gene expression in oral squamous cell carcinoma using machine learning applications

Abstract Background Recently, the possibility of tumour classification based on genetic data has been investigated. However, datasets are difficult to handle because their massive size and complexity manipulation. In present study, we examined diagnostic performance machine learning applications using imaging-based classifications oral squamous cell carcinoma (OSCC) gene sets. Methods RNA sequencing from SCC tissues various sites, including oral, non-oral head neck, oesophageal, cervical regions, were downloaded The Cancer Genome Atlas (TCGA). feature genes extracted through a convolutional neural network (CNN) learning, each analysis was compared. Results ability classify OSCC tumours excellent. tool exhibited poorer in discriminating histopathologically dissimilar cancers derived same type tissue than differentiating histopathologic with different origins, revealing that differential expression pattern is more important factor features for cancer types. Conclusion CNN-based model visualisation methods useful correctly categorising OSCC. showed differentially expressed multiwise comparisons types SCCs, such as KCNA10 , FOSL2 PRDM16 leader pairwise FGF20 DLC1 ZNF705D .