Evolutionary artificial intelligence based peptide discoveries for effective Covid-19 therapeutics.
Biochim Biophys Acta Mol Basis Dis
; 1867(1): 165978, 2021 01 01.
Article
in English
| MEDLINE | ID: covidwho-1023476
ABSTRACT
An epidemic caused by COVID-19 in China turned into pandemic within a short duration affecting countries worldwide. Researchers and companies around the world are working on all the possible strategies to develop a curative or preventive strategy for the same, which includes vaccine development, drug repurposing, plasma therapy, and drug discovery based on Artificial intelligence. Therapeutic approaches based on Computational biology and Machine-learning algorithms are specially considered, with a view that these could provide a fast and accurate outcome in the present scenario. As an effort towards developing possible therapeutics for COVID-19, we have used machine-learning algorithms for the generation of alignment kernels from diverse viral sequences of Covid-19 reported from India, China, Italy and USA. Using these diverse sequences we have identified the conserved motifs and subsequently a peptide library was designed against them. Of these, 4 peptides have shown strong binding affinity against the main protease of SARS-CoV-2 (Mpro) and also maintained their stability and specificity under physiological conditions as observed through MD Simulations. Our data suggest that these evolutionary peptides against COVID-19 if found effective may provide cross-protection against diverse Covid-19 variants.
Keywords
Full text:
Available
Collection:
International databases
Database:
MEDLINE
Main subject:
Peptides
/
Artificial Intelligence
/
COVID-19 Drug Treatment
Type of study:
Experimental Studies
/
Prognostic study
/
Randomized controlled trials
Topics:
Vaccines
/
Variants
Limits:
Humans
Language:
English
Journal:
Biochim Biophys Acta Mol Basis Dis
Year:
2021
Document Type:
Article
Affiliation country:
J.bbadis.2020.165978
Similar
MEDLINE
...
LILACS
LIS