A PIM-1 Kinase Inhibitor Docking Optimization Study Based on Logistic Regression Models and Interaction Analysis.

Area of Science:

• Medicinal Chemistry
• Computational Biology
• Oncology

Background:

• PIM-1 kinase is a key target in multiple cancers, including prostate, breast, and blood cancers.
• Accelerating the discovery of PIM-1 kinase inhibitors is crucial for developing targeted oncological drugs.
• Computer-aided screening offers a promising avenue for novel therapeutic discovery but faces limitations.

Purpose of the Study:

• To develop an improved post-processing method for molecular docking results in virtual screening.
• To enhance the accuracy of identifying PIM-1 kinase inhibitors by integrating binding modes and in vitro data.
• To refine computational strategies for hit discovery in kinase inhibitor research.

Main Methods:

• A novel post-processing method for molecular docking scores and binding affinities was developed.
• The method incorporates known inhibitor binding modes and available in vitro data.
• Logistic regression models were trained using binding energies and key amino acid residue interactions to predict inhibitory activity.

Main Results:

• The docking protocol effectively distinguished known PIM-1 kinase inhibitors from decoy molecules.
• Binding energies alone were insufficient for accurate prediction; incorporating binding modes and interactions improved results.
• The logistic regression model achieved 80.9% true positive and 81.4% true negative rates in predicting PIM-1 inhibitory activity.

Conclusions:

• The proposed post-processing approach enhances the success rate of hit discovery in virtual screening campaigns.
• This method offers a more reliable strategy for identifying potential PIM-1 kinase inhibitors.
• The approach can be applied to large-scale molecular docking for accelerated therapeutic development.