Reflections: What Really Matters  |  Teacher Insights: New Harvard Online course course prepares professionals for a data-driven world  |  Parent Interventions: Research shows lullabies in any language relax babies  |  International Edu News: 'Plastic bags could be 'eco-friendlier' than paper and cotton bags'  |  Leadership Instincts: Start-up with plastic waste recycling solution wins top prize at ideasinc 2020  |  International Edu News: Frailty, old age and comorbidity main predictors of death from Covid-19  |  Leadership Instincts: Cyber centre to reduce digital harm  |  Policy Indications: New funding to improve water security for 10 million people in Africa and Asia  |  International Edu News: UCL hosts global conference on UN Sustainable Development Goals  |  International Edu News: Medium-term impact of COVID-19 revealed in new study  |  International Edu News: Extremely rapid diagnostic test for Covid-19  |  Teacher Insights: Cambridge University Press to join with Cambridge Assessment  |  National Edu News: Minister inaugurates new Diamond Jubilee Lecture Hall Complex of NIT Jamshedpur  |  Education Information: CSIR partnered clinical trials website “CUReD” on Repurposed Drugs for Covid- 19  |  Teacher Insights: The 6th India International Science Festival to be held in Virtual format  |  
September 15, 2020 Tuesday 12:39:22 PM IST

New Model Predicts Covid Mutations

Science Innovations

A novel machine learning model developed by scientists at Michigan State University have identified more mutations of the SARS-CoV-2 genome that have made the virus more infectious.  More than 20,000 viral genome samples were analysed for mutations to the spike protein. It was found that five of the six known virus subtypes are now more infectious. Knowledge of mutations is important to understand the transmission, diagnostics, prevention and treatment of the disease. The binding affinity of the spike protein during the initial stage of infection is vital for the onset of the disease. Increased binding activity was observed by analysing more than 8000 protein interaction records. The binding activity was seen to have increased in six known subtypes. The model predicts that multiple residues on the receptor-binding motif- a component area of the receptor-binding domain -- have high changes to mutate into more infectious Covid-9 strains.