A protein kinase is a kinase enzyme that modifies other proteins by chemically adding phosphate groups to them (phosphorylation). These protein kinase are critical component of cellular signal transduction and involved in several diseases as cancer and inflammation.
The chemical activity of a kinase involves removing a phosphate group from ATP and covalently attaching it to one of three amino acids that have a free hydroxyl group. Most kinases act on both serine and threonine, others act on tyrosine, and a number (dual specificity kinases) act on all three.
Chemical kinase inhibitors are class of therapeutics molecules which may bind with protein kinases. The wast majority of of these compounds binds to the ATP binding site, and because all protein kinases (518) of human are highly conserved at this site. Hence there is great potential of cross-reactivity. Compounds must be tested with other protein kinases for specificity and off-target interactions.
Designing a specific molecule inhibitor for specific protein kinase is always area of interest.Unfortulately researchers can't developed any specific molecule for any kinase. Now attension sifted towards design a molecule which may bind with intended target with better binding comparison to rest other kinases.
DMKPred is a SVM based webserver for the prediction of binding of chemical molecules with specific kinases. Here user can paste or upload the molecular descriptors off chemical molecues and our prediction method calculate the binding of these chemical with protein kinases.
During our study we calculate the molecular descriptors using online software PreADMET. We select top 5 molecular descriptors of chemical molecule for all protein kinases model, which have highest correlation with Kd. So user must read the list of descriptors for each protein kinases before runnig this software.