Protein kinase inhibitors' classification using K-Nearest neighbor algorithm.

Area of Science:

• Biochemistry and Molecular Biology
• Medicinal Chemistry
• Computational Drug Design

Background:

• Protein kinases regulate biological activities via phosphorylation, making them key targets in cancer therapy.
• Small molecule inhibitors targeting protein kinases are crucial for cancer treatment.
• Quantitative Structure-Activity Relationship (QSAR) modeling offers an economical approach for drug design.

Purpose of the Study:

• To develop and evaluate a hybrid QSAR model for distinguishing active from inactive protein kinase inhibitors.
• To optimize computational drug design strategies for identifying potential cancer therapeutics.

Main Methods:

• Utilized a hybrid QSAR model incorporating genetic algorithms for dimensional reduction.
• Employed the K-Nearest Neighbors method for classification of inhibitor activity.
• Evaluated model performance using Support Vector Machine and Naïve Bayesian algorithms.

Main Results:

• The proposed hybrid QSAR model demonstrated superior performance in distinguishing active inhibitors.
• Genetic algorithms effectively reduced dimensionality for QSAR model development.
• K-Nearest Neighbors provided robust classification of small molecule inhibitors.

Conclusions:

• The developed hybrid QSAR model is a powerful and cost-effective tool for identifying potential protein kinase inhibitors.
• This approach significantly advances computational drug design for targeted cancer therapies.
• The model's superiority suggests its utility in accelerating the discovery of novel anti-cancer agents.