Deep Learning Analysis of Single-Cell Data Reveals Unique Genetic Features Of COVID-19 Severities-Reference-Cited by-同舟云学术

Deep Learning Analysis of Single-Cell Data Reveals Unique Genetic Features Of COVID-19 Severities

Published:2023-06-12 Issue: Volume: Page:
ISSN:
Container-title:
language:
Short-container-title:

Author:

Schonfeld Elan¹^ORCID

Affiliation:

1. Columbia Univeristy

Abstract

Abstract COVID-19 was declared by the World Health Organization in 2020 to be a pandemic. Analysis of COVID-19 related genetic pathways allows for a better understanding of the possible effects and sequelae of the disease. Using 6178 scRNA sequenced human cells, having a status of control/mild/severe COVID-19 disease status, differential expression of genes and pathways was analyzed. Using Gene Set Enrichment Analysis (GSEA), mild COVID-19 was found to over-express the Influenza Pathway. In order to identify genes important in COVID-19 severity, a deep learning classifier was trained. Classifiers were repeatedly trained for this task using 10 randomly selected genes from the total number of 18,958 genes. The highest performing classifier (AUC=0.748) was trained using: AC008626.1, SGO1, RHOBTB2, RBM41, NDUFAF4P1, COX5A, ZDHHC17, STX11, IPP, NUDT5 genes. These results further illustrate the other factors contributing to mild versus severe COVID-19, as well as evidence of potential misdiagnosis or overlapping pathway effects of Influenza and COVID-19.

Publisher

Research Square Platform LLC

Reference23 articles.

1. Lv M, Luo X, Estill J,etal.Coronavirusdisease(covid-19):Ascopingreview.Eurosurveillance.https://www.ncbi.nlm.nih.gov/pmc/articles/PMC7175649/.PublishedApril2020.AccessedNovember16,2021.

2. Chowdhury MA, Hossain N, Kashem MA, Shahid MA, Alam A.ImmuneresponseinCOVID-19:AReview.Journalofinfectionandpublichealth.https://www.ncbi.nlm.nih.gov/pmc/articles/PMC7359800/.PublishedNovember2020.AccessedNovember16,2021.

3. Zhang Y, Wang D, Peng M,etal.Single-cellRNAsequencingincancerresearch.JournalofExperimental&ClinicalCancerResearch.https://jeccr.biomedcentral.com/articles/10.1186/s13046-021-01874-1#:~:text=scRNA%2Dseq%20shows%20that%20one,tumor%20formation%2C%20and%20cancer%20relapse.PublishedMarch1,2021.AccessedNovember16,2021.

4. Ramirez Flores RO, Lanzer JD, Holland CH,etal.Consensustranscriptionallandscapeofhumanend-stageheartfailure.JournaloftheAmericanHeartAssociation.https://www.ahajournals.org/doi/10.1161/JAHA.120.019667.PublishedMarch31,2021.AccessedNovember18,2021.

5. Higdon R.Genesetandproteinsetexpressionanalysis|SpringerLink.EncyclopediaofSystemsBiology.https://link.springer.com/referenceworkentry/10.1007%2F978-1-4419-9863-7_1209.AccessedNovember16,2021.