Abstract
Hepatocellular carcinoma (HCC) is one major cause of cancer-related mortality around the world. However, at advanced stages of HCC, systematic treatment options are currently limited. As a result, new pharmacological targets must be discovered regularly, and then tailored medicines against HCC must be developed. In this research, we used biomarkers of HCC to collect the protein interaction network related to HCC. Initially, DC (Degree Centrality) was employed to assess the importance of each protein. Then an improved Graph Coloring algorithm was used to rank the target proteins according to the interaction with the primary target protein after assessing the top ranked proteins related to HCC. Finally, physio-chemical proteins are used to evaluate the outcome of the top ranked proteins. The proposed graph theory and machine learning techniques have been compared with six existing methods. In the proposed approach, 16 proteins have been identified as potential therapeutic drug targets for Hepatic Carcinoma. It is observable that the proposed method gives remarkable performance than the existing centrality measures in terms of Accuracy, Precision, Recall, Sensitivity, Specificity and F-measure.