De novo identification of substrates for enzymes is typically done experimentally. However, the approach takes time and requires a recombinant protein to test the predictions. While for many proteins it is a routine, for others it becomes problematic due to expression problems, solubility/aggregation of the purified biomolecule and time constraints. Therefore, in silico approaches which offer a chance to bypass the tedium are very desired. In the recent work by Swietnicki and Brzozowska https://doi.org/10.1016/j.jmgm.2019.07.002, the authors used computational tools to identify putative binding site for a sugar substrate. A closely related carbohydrate was not the substrate and the in silico tools were used to differentiate between the 2 molecules. Analysis of binding energies and retention on the molecule during molecular dynamics simulation showed that the difference between molecules was most likely attributed to the residence time of carbohydrates on the protein. The authors went a step further and analyzed the binding site of the enzyme to predict putative preferences for the enzyme and elucidate its catalytic mechanism.
The methodology used by the authors has been in use in the commercial sector for many years for a different work. Application of the approach to bacteriophages may have a future as an alternative to experimental screening of bacteria against a panel of phages. If the pre-screening is performed in silico based on DNA sequence of a pathogen and a constructed in silico library of bacteriophage proteins, selection of phages could be substantially shortened. One could also imagine a personalized bacteriophage therapy for infectious, and not only, diseases in humans as a future medical approach to treating bacterial infections. For the drug-resistant pathogens, bacteriophage therapy could be personalized as any other branch of medicine.
The work was performed at the Institute of Immunology and Experimental Therapy of the the Polish Academy of Science in Wroclaw, PL and guided by Dr. Swietnicki. Computational resources were from the on-site IT Department and the software was purchased from Schrodinger LLC (www.schrodinger.com).