The Analysis Of SNPs In Successful Drug Targets

There are millions of SNPs in human genome, and many of them can transfer to proteins followed by genes expression. Therefore, they will directly or indirectly lead to a decrease of drug-target interactions, and then confer drug resistance. The HapMap database contains thousands of SNP information coming from different human species. As different human species have different SNP frequency, it will be the same case to their drug targets, hence different human species may confer difference in drug resistance, that is to say, different drugs may be suitable for different human species. Although many troubles have occurred in administration to different human species, up to date, researchers have not systemically studied the drug resistance between targets and their drugs.By combining with information of human SNPs in HapMap, we analyzed69successful drug targets, which have at least one launched drugs for each target and a total of158SNPs, and their drugs in Therapeutic Targets Database (TTD). The results show that for most drug targets, the SNP sites are far from their active pocket (Distance>15A,146SNPs, occupied92.4%fo the total158SNPs, and57targets, occupied82.6%of the total69targets); a small number of targets have SNPs near to the active pockets (6A<Distance<15A,12SNPs, occupied7.6%fo the total158SNPs, and12targets, occupied17.4%of the total69targets); and there are no targets with the SNP sites located in their active pockets (Distance<6A,0,0%). As the SNPs located near the active pocket may affect the drug-target interactions, we further the analysis of these drug targets whose SNPs are close to the active pockets using many methods, such as molecular dynamics, free energy calculation (MM/GBSA), residue decomposition, principal component analysis.In order to evaluating the drug-target interactions accurately, we firstly chose a mutated drug target--ALK (Anaplastic Lymphoma Kinase) with a SNP site C1156Y, and this mutated drug target has been proved to confer drug resistance by many experiments. Then we calculated the free energy between the protein and drug in the wild ALK (WT) and the mutated ALK (C1156Y) by molecular dynamics and MM/GBSA free energy calculation. The result shows that MM/GBSA is effective in ranking the binding affinities of the drugs in WT and C1156Y (in the WT, the enthalpy is-37.67kcal/mol, while-33.61kcal/mol in the C1156Y respectively). After that, we studied the targets whose SNPs are close to the active pocket using the approaches above, and the results show that three drug targets may confer drug resistance (Cytochrome P4503A4, Plasminogen, Peroxisome Proliferator Activated Receptor Alpha). Our research may provide some information for the option of administration to different human species and personal drug design.