Protein-protein interaction prediction from primary sequences using supervised machine learning algorithm

Protein-Protein Interactions (PPI) study is significant to comprehending cellular biological functions. Though there are different experimental techniques to predict PPIs, detecting PPIs in the lab is costly and time-consuming. Nowadays, high throughput approaches and large-scale biological techniques have achieved incredible growth. These large-scale techniques experience false positive and false negative predictions. As a result, there is a need to devise a computational technique for estimating PPI pairs, which complements laboratory techniques and offers an inexpensive way to find the interactions between proteins. Although much advancement has been achieved for PPI prediction still there is a requirement for a much more effective approach to predict PPI from protein sequences. The proposed model gives 93% accuracy, 92.9% sensitivity, 92.6% precision, 92.5% specificity, and 92.7% f1-score. The results indicate that our proposed model outperforms various predictors for PPI prediction.