Enhanced ranking of PknB Inhibitors using data fusion methods.

Area of Science:

• Computational chemistry
• Drug discovery
• Molecular biology

Background:

• Mycobacterium tuberculosis has 11 serine-threonine protein kinases (STPKs) regulating vital cellular processes.
• PknB is essential for mycobacterial growth, phosphorylating peptidoglycan biosynthesis substrates.
• High-affinity PknB inhibitors are actively sought for tuberculosis treatment.

Purpose of the Study:

• To evaluate data fusion algorithms for enhancing virtual screening of PknB inhibitors.
• To identify novel PknB inhibitors using combined structure- and ligand-based approaches.

Main Methods:

• Employed structure-based and ligand-based virtual screening.
• Applied data fusion techniques including sum rank, sum score, and reciprocal rank.
• Screened the Asinex database using the reciprocal rank algorithm.

Main Results:

• Data fusion significantly improves the ranking of active compounds compared to individual methods.
• The reciprocal rank algorithm outperformed other fusion methods and single approaches.
• Identified 45 candidate compounds for PknB inhibition via fused virtual screening.

Conclusions:

• Data fusion effectively integrates diverse prediction methods for drug-target interactions.
• Combined approaches demonstrate superior performance in ranking active compounds.
• Fused results offer a robust strategy for selecting candidates for biological screening.