Predictive Analytics to Combat with COVID-19 using Genome Sequencing

Abstract

The coronavirus outbreak (nCoV) has created an alarming situation at international level. COVID-19 pandemic is now continuing to develop and researchers across the world are working to combat with COVID-19 & to reduce and prevent its spread. Pathological tests are used to detect whether patient is corona positive or not but they are time consuming & also number of test kits & centers are limited. Surprising thing is 60% patients may not have any symptoms but still they are SARS-CoV-2 positive & may work as silent carrier causing outbreak of this disease. And a reason to worry is that yet there is no vaccine & exact drug for this disease.

Clinical data related to patient’s medical history, geographical location & symptoms can be analyzed using machine learning techniques to predict clinical outcomes. But as symptoms vary from person-to-person & are common in other diseases like flu, swine flu & pneumonia therefore accuracy of clinical outcomes is questionable. To tackle this problem, this paper proposes a technique to detect presence of SARS-CoV-2 in a person using genome sequencing of virus. This will accurately help to detect presence of virus, to develop targeted therapies & vaccines and also to learn how patient will respond to drugs. This paper summarizes primary survey of existing work & discusses a methodology to detect. Thus proposed methodology can be used by doctors & practitioners as a tool for making ease in decision making.

Keywords: COVID-19; machine learning; predictive analytics; genome sequencing SARS-CoV-2; India pandemic

JEL Classification: I1, I14, I10