The linkage between genetic variants and human diseases is well-established fact providing promising direction for early detection of disease risk and appropriate treatment. Essential component in utilizing the linkage is to reveal molecular effects that are associated with the disease-causing variants since this will provide efficient way of structure-based drug discovery. In this talk, we focus on a specific class of disease-causing molecular effects associated with alteration of protein-DNA binding, while do not affecting wild type protein stability and dynamics. Thus, the protein variant retains the wild type characteristics and is expected to be intact, but its binding affinity or specificity to cognate DNA are altered. This offers the possibility of developing small molecules as potential drugs. Their binding to mutant proteins and cognate DNA restores the wild type binding characteristics and thus eliminates disease-causing effect. As a particular case, we demonstrate that most-frequently occurring missense mutations linked with Rett syndrome indeed do not affect the corresponding protein (the methyl binding domain (MBD) of MeCP2 protein), but alter MBD interactions with DNA. In silico screening identified set of small molecules which binding to MBD-DNA mitigates the effect of disease-causing mutations. At the end, we outline the computational and mathematical challenges associated with small molecules targeting protein-DNA interactions.