Calculation of protein mutability landscape and thereonforecasting evolutionary pathways: Neuraminidase of H1n1virus as a case study

Viruses and bacteria have been among the most harmful agents for human health in history. Many lives have been saved with the discovery of antibiotics and antiviral drugs. However, the rapid emergence of resistant strains became an ever-increasing health concern since the 1960s. These resistant strains are capable of inactivating the drug efficacy and survive in infected cells successfully. Therefore, it is very important to analyze evolutionary pathways of viruses and understand their susceptibility and robustness to mutation with generating mutability landscape. Thus, predicting future mutant strains with the help of mutability probabilities is a potential to discover new drug candidates before emerging as a threat for humans. Despite decades of research, forecasting evolutionary pathways remains extremely challenging due to lack of both available data and appropriate methods. So far, amino acid frequency and substitution matrixes are the most widely used parameters in calculation of protein mutability. Here, we developed a model to predict the strains that may appraise in the future. The swine flu H1N1, caused many deaths, is used as a case study. We generated the mutability landscape for neuraminidase protein in swine flu according to our mutability probabilities. Thus, we addressed the location of conserved and non-conserved residues in neuraminidase. With using amino asid frequencies, mutability landscape and mutation rate of neuraminidase, we have forecast the sequences. This prediction model may lead to obtain more accurate prediction in the future and allow us to design novel drugs in advance.