In the past three years, the COVID-19 pandemic has taken a major toll, not only on human lives, but also on economic and socio-political situations in countries around the world. Besides coming up with vaccines within a short time, Indian researchers have been tackling other types of approaches to counter the pandemic. Here is a brief overview of Indian contributions from January 15 to 31, 2022.
Monitoring the malady
Screening and isolating infected individuals are primary tasks for flattening the infection curve. Infrared thermal scanners are now common, but involve time and labour and fail to cover larger regions.
Scientists from KIIT University, Bhubaneswar and SRM University, Andhra Pradesh have now come up with an easier method . Besides fever, continuous dry cough is a major symptom of COVID infection. Why not create an automated system to monitor the symptoms?
To detect the spread of the virus, the researchers developed a system based on the Internet of Things in collaboration with researchers from Vietnam. They created a device with a sensor to record persistent cough and to raise alerts. The device also has a temperature sensor for continuously measuring body temperature and an ultrasonic sensor to detect people who violate the six foot physical distance. The team designed their prototype such that it can be easily mounted over clothing.
They tested their prototype in an aluminium ingots factory, and found it successful.
To cope with the pandemic, their Smart COVID-Shield is a good option for workplaces and educational institutions, claim the researchers.
Each wave of the pandemic tends to put pressure on the health services. It becomes difficult for doctors to diagnose and distinguish COVID-19 from other respiratory infections.
RT-PCR tests take time. But taking digital X-ray images is easy. To automatically distinguish COVID-19 from X-ray images, Arun Kumar and Rajendra Prasad from the SRM Institute of Science and Technology, Delhi have now come up with a deep-learning model .
They first tested models based on convolution neural networks and deep neural networks.
The deep neural network had about 85% accuracy whereas convolution neural network algorithms achieved about 95%.
The researchers say that the convolution neural network model can replace even RT-PCR
which is time-consuming and expensive.
Another machine learning method which has 100% accuracy was proposed by Nivetha and Hannah from Periyar University, Salem . But, to get this level of accuracy, they used CT scan images – slightly costlier than X-ray images.
Their approach was based on the Neighbourhood Rough Neural Network – a deep-learning method to automatically distinguish COVID-19 from other respiratory diseases.
The method has potential to be deployed for the automatic detection of other diseases from CT scan images.
Messing with medicines
The first two years of the pandemic saw doctors messing with medicines. Many different medicines which did not have any effect were tried on unsuspecting patients – some even had severe side effects.
The days of trial and error in drug discovery will soon come to an end. There are now computational tools to identify differentially expressed genes in pathology and molecular simulation and docking to identify the potential chemical compounds in the existing drugs library – a faster and much safer option for drug discovery.
Recently, scientists from the Delhi University, Jamia Millia Islamia, JNU and AIIMS, New Delhi performed an in-silico analysis to identify potential drug targets for COVID 19 treatment .
Inflammation is the main pathology that troubles Covid patients. So they conducted extensive analyses of inflammatory pathways and identified the ARHGEF1 gene associated with COVID‐19. ARHGEF1 gene-mediated regulation of immunity was found to be a controlling factor in chronic inflammation in infected patients.
Using molecular simulation and docking studies, the researchers also found a few possible drugs for acting against this target, drugs that could be used in combination with other COVID-19 medications.
However, extensive in-vitro, in-vivo and clinical trials are required to confirm their efficacy and safety.
 Hrudaya Kumar Tripathy, Sushruta Mishra, Shubham Suman, Anand Nayyar & Kshira Sagar Sahoo, Smart COVID-shield: an IoT driven reliable and automated prototype model for COVID-19 symptoms tracking, Computing (2022); DOI: 10.1007/s00607-021-01039-0
 S. Nivetha & H. Hannah Inbarani, Neighborhood Rough Neural Network Approach for COVID-19 Image Classification, Neural Processing Letters (2022); DOI: 10.1007/s11063-021-10712-6
 Arun Kumar, Rajendra Prasad Mahapatra, Detection and diagnosis of COVID-19 infection in lungs images using deep learning techniques, International journal of Imaging Systems and Technology (2022); DOI: 10.1002/ima.22697
 Prakash Jha, Prithvi Singh, Shweta Arora, Armiya Sultan, Arnab Nayek, Kalaiarasan Ponnusamy, Mansoor Ali Syed, Ravins Dohare, Madhu Chopra, Integrative multiomics and in silico analysis revealed the role of ARHGEF1 and its screened antagonist in mild and severe COVID-19 patients, Journal of Cellular Biochemistry (2022); DOI: 10.1002/jcb.30213