Enhanced Parkinson’s Disease Genes Prediction

Identifying genes associated with Parkinson’s disease (PD) is an interesting research area that aids in the disease’s diagnosis and treatment. Most PD approaches were created only to find protein-coding genes and ignore long non-coding (lncRNA) genes, which are crucial for biological functions, disease transformation, and development. In this paper, the proposed system is developed to predict protein and lncRNA genes associated with PD, which consists of four steps. First, preprocessing step is to represent the genes as DNA FASTA sequences. Second, feature extraction methods are used to extract and discover the most discriminative features of the biological sequences, and the extracted features are fused. Third, the Adaboost technique is used as feature selection to reduce these fused feature vectors’ dimensionality. Finally, the gradient-boosted decision tree algorithm is used for diagnosing various test cases. The proposed prediction system achieves the best result compared with state-of-arts studies with an area under the curve (AUC) equal to 86.7%.