Machine Learning Integrated Drug Discovery Process Investigating Possible Drug Molecule against COVID-19 Disease: A Review

Pandemic is an outbreak of infectious disease that results in economic damage and loss of life on a global scale. Recently the world has suffered due to the pandemic known as COVID-19 disease caused due to the virus SARS-CoV-2, named by World Health Organization (WHO). The COVID-19 infected cases on global scale reported by the WHO was estimated to be 771,820,937. The global spread of COVID-19 is majorly due to the higher rates of mutation observed in the S1 spike protein of SARS-CoV-2 virus, and it is found to be one of the major difficulties in developing novel drugs and vaccines in a short period of time. Till today, drugs and vaccines developed against COVID-19 functioned as first aid treatment option instead of eliminating the disease completely. To apprehend, there is an indefinite need in developing robust Machine Learning (ML) algorithm for designing effective drugs and vaccines in a fleeting period to fight against viral outbreaks such as COVID-19 disease. This study signifies research gaps notedly, limited access to virtual screening, lack of general ML pipeline for drug discovery, non-availability of experimental data in designing ML algorithms and insufficient molecular docking results; were to be addressed developing integrated ML architecture pipeline. This study concludes that, to overcome the challenges discussed, there is a certain need in developing robust ML architectural layout consisting different ML algorithms integrated to apply for, the process of discovering novel drugs against viral protein in order to fight against pandemics such as COVID-19 disease.