Genome-wide Machine Learning Analysis of Anosmia and Ageusia with COVID-19

Lucas Pietan; Elizabeth Phillippi; Marcelo Melo; Hatem El-Shanti; Brian J Smith; Benjamin Darbro; Terry Braun; Thomas Casavant

doi:10.1101/2024.12.04.24318493

Back

Genome-wide Machine Learning Analysis of Anosmia and Ageusia with COVID-19

Preprint

Open access

Genome-wide Machine Learning Analysis of Anosmia and Ageusia with COVID-19

Lucas Pietan, Elizabeth Phillippi, Marcelo Melo, Hatem El-Shanti, Brian J Smith, Benjamin Darbro, Terry Braun and Thomas Casavant

medRxiv : the preprint server for health sciences

Cold Spring Harbor Laboratory

12/05/2024

DOI: 10.1101/2024.12.04.24318493

PMCID: PMC11643161

PMID: 39677430

Files and links (1)

url

https://doi.org/10.1101/2024.12.04.24318493View

Preprint (Author's original)This preprint has not been evaluated by subject experts through peer review. Preprints may undergo extensive changes and/or become peer-reviewed journal articles. Open Access

Abstract

The COVID-19 pandemic has caused substantial worldwide disruptions in health, economy, and society, manifesting symptoms such as loss of smell (anosmia) and loss of taste (ageusia), that can result in prolonged sensory impairment. Establishing the host genetic etiology of anosmia and ageusia in COVID-19 will aid in the overall understanding of the sensorineural aspect of the disease and contribute to possible treatments or cures. By using human genome sequencing data from the University of Iowa (UI) COVID-19 cohort (N=187) and the National Institute of Health All of Us (AoU) Research Program COVID-19 cohort (N=947), we investigated the genetics of anosmia and/or ageusia by employing feature selection techniques to construct a novel variant and gene prioritization pipeline, utilizing machine learning methods for the classification of patients. Models were assessed using a permutation-based variable importance (PVI) strategy for final prioritization of candidate variants and genes. The highest held-out test set area under the receiver operating characteristic (AUROC) curve for models and datasets from the UI cohort was 0.735 and 0.798 for the variant and gene analysis respectively and for the AoU cohort was 0.687 for the variant analysis. Our analysis prioritized several novel and known candidate host genetic factors involved in immune response, neuronal signaling, and calcium signaling supporting previously proposed hypotheses for anosmia/ageusia in COVID-19.

Details

Title: Subtitle: Genome-wide Machine Learning Analysis of Anosmia and Ageusia with COVID-19
Creators: Lucas Pietan
Elizabeth Phillippi
Marcelo Melo
Hatem El-Shanti
Brian J Smith
Benjamin Darbro
Terry Braun
Thomas Casavant
Resource Type: Preprint
Publication Details: medRxiv : the preprint server for health sciences
DOI: 10.1101/2024.12.04.24318493
PMID: 39677430
PMCID: PMC11643161
Publisher: Cold Spring Harbor Laboratory; United States
Language: English
Date posted: 12/05/2024
Academic Unit: Roy J. Carver Department of Biomedical Engineering; Electrical and Computer Engineering; Stead Family Department of Pediatrics; Biostatistics; Medical Genetics and Genomics; Holden Comprehensive Cancer Center
Record Identifier: 9984757687702771

Metrics

68 Record Views