プロテオミクスとバイオインフォマティクスのジャーナル
オープンアクセス

概要

A New Bioinformatics Method based on Chou’s PseAAC Concept for Prediction of Carcinogenicity of Several Proteins of HPV

Hassan Mohabatkar, Komail Amini, Parisa Rabiei, Kamran Mansouri

Objective: Today, Human Papilloma Viruses (HPV) is considered as second most frequent carcinogens. It has been distinguished that structural proteins of HPVs are involved in their carcinogenicity. The aim of this work was analysis of the HPV proteins and classifying them into two categories, low and high-risk, based on their physiochemical properties and Chou’s Pseudo Amino Acid concept.

Materials and Methods: Initially, sequences of 69 proteins belonging to high-risk viruses and 107 proteins belonging to low-risk viruses were collected from NCBI database and Uniprot. Then the PseAAC server was used to analyze these sequences. In the next step, the information obtained from this server was analyzed software. The KNN algorithm was used to classify the datasets. Furthermore, 10 and 6 fold-cross validation test was applied on the classifier to evaluate our prediction method.

Results: Specificity, sensitivity, and accuracy of our proposed method were achieved 87.64%, 87.22% and 87.39% in the treated dataset and 91.11%, 82.60% and 86.66% in the tested dataset, respectively.

Conclusion: In this study, we developed a sequence-based method for predicting the carcinogenicity potent of HPV proteins and dividing them into high risk and low-risk categories.