Back

Discovery of dual inhibitors of KPC-3 and KPC-15 of Klebsiella pneumoniae - a ligand and structure based virtual screening study

Sharma, S.; Kumar-M, P.; Yadav, R.; Gupta, N.

2019-12-23 bioinformatics
10.1101/2019.12.23.886671 bioRxiv
Show abstract

BackgroundThe development of carbapenem resistance against Klebsiella pneumoniae is a situation of grave concern and requires urgent attention. Among the KPC produced by K.pneumoniae, KPC-3, and KPC-15, play a significant role in the development of resistance to carbapenem. Materials and methodsThe binding sites of KPC-3 and KPC-15 were predicted by the COACH server. Drug-like ligands from ZINC were then screened by ligand-based drug screening (LBVS) by keeping Relebactam as a template. The top 50,000 selected ligands were then screened by structure-based virtual screening using idock. For keeping an account of the dual inhibitors stability in complex with KPC-3 and KPC-15, MDS were carried out for each complex. ResultsBased on consensus weighted ranks, the top 3 ligands with the dual inhibitory property are ZINC76060350 (consensus weighted rank - 1.5), ZINC05528590 (2), ZINC72290395 (3.5). All the top 3 dual inhibitors have a reasonable probability of passing through the blood-brain barrier. The RDKit and Morgan fingerprint scores between Relebactam and the top three ligands were 0.24, 0.22, 0.23, and 0.26, 0.19, 0.25, respectively (showing only 20% similarity). The MD simulation result revealed good binding stability of ligand ZINC05528590 with both KPC-3 and KPC-15, whereas ligand ZINC76060350 showed good binding stability to KPC-3. ConclusionThe ligand ZINC05528590 could be taken forward to develop a new drug against a multi-resistant- Klebsiella pneumoniae infection. At the same time, ZINC76060350 can be considered to develop a new drug against KPC-15 resistant Klebsiella pneumoniae.

Matching journals

The top 8 journals account for 50% of the predicted probability mass.

50% of probability mass above

"Similar papers" are the closest papers from that journal in the model's embedding space. They show what the match is built on, but the ranking comes mostly from a classifier over the whole training set, not from these examples alone.