COVID-19-Mutation-Rate-Analysis-using-Recurrent-Neural-Network-based-LSTM-model

The new coronavirus SARS-CoV-2, also known as COVID-19, has caused a global pandemic. This contagious RNA virus threatened the entire world. More than 7.9 million people have been infected as of June 15, 2021 with 432k deaths. This virus is capable of causing mutations in humans. Understanding the evolution of this virus and determining the risk of the emerging infectious disease requires accurate mutation rate determination. This research investigates the mutation rate of the entire genome sequence based on patient data. The obtained data is analysed independently to detect nucleotide and codon mutations. Furthermore, the mutation rate analysis shows that a large number of Thymine (T) and Adenine (A) nucleotides are mutated when compared to other nucleotides, although the mutation rate of codons is not frequent when compared to nucleotides. A recurrent neural network (RNN) based Long Short-Term Memory (LSTM) model was used to predict the future mutation rate. This model gives very less error it is around 1.03 in both the training and testing sets. The mutation of the 83rd patient has been predicted by this model.