Comprehensive discovery of markers effective in disease severity in the genome of Iranian patients with covid-۱۹

This article examines the use of artificial intelligence algorithm statistical techniques forautomatic identification of biomarkers with the aim of prognosticating the severity of infection inpatients with COVID-۱۹. The operation can be applied to other viruses as well. In this research, a groupof people (۱۱۰ people) were selected for the study and they were classified into two groups of patientswith high and low severity, and the evolution-based harmony search algorithm was used to extractpatterns from the virus genome and also, a multi-objective approach was used to investigate therelationship between these patterns and the severity of the disease. In the next part of the study, thetargets of unique miRNAs related to the genome of samples of two groups are extracted and consideredas a marker to predict the severity of infection. Also, taking into account the effect of different aminoacids on increasing the potential of their interactions, a method for grouping and weighting wasproposed, which has improved the search capability of the algorithm. The results of this model were inmany cases consistent with previously identified markers determined through experiments. Theobtained results show that there is a significant difference in the mutation rate in the candidate regionsin patients with high and low severity. Also, the relationship between mutations and the age of thepatients was investigated and a significant correlation was observed between the frequency of mutationsin the designated areas and the age of the patients.

کلیدواژه ها:

artificial intelligence algorithm ، mutations ، biomarkers of molecular diagnosis