# Metaproteomics analysis of respiratory tract samples from COVID-19 infected patients

# Live Resources

usegalaxy.eu
data library data library view view
Input data Input data view view
Result history Result history view view
workflow workflow run run

# Description

Cardozo et al (opens new window) collected bottom-up mass spectrometry (MS) data on respiratory tract samples from ten COVID-19 positive patient samples. Data-dependent acquisition MS spectra were acquired using hybrid quadrupole-Orbitrap tandem mass spectrometry. The MS data was used to generate a spectral library of targeted COVID-19 peptides for targeted MS assay for clinical samples. We were interested in exploring the possibility of presence of microorganisms in the clinical samples. Peter Thuy-Boun from Wolan Lab at the Scripps Institute searched the five RAW files (pools 18, 34, 38, 47 and 51) using COMPIL 2.0 (opens new window) against a comprehensive 113 million protein sequence database. The detected peptides identified were subjected to Unipept 4.3 analysis to detect taxonomic information about microorganisms present in the sample. A list of clinically significant genera/species was used to generate a protein FASTA database within the Galaxy workflow (opens new window). The generated protein database along with the RAW files and COVID-19 protein database was used as inputs for a Galaxy workflow to a) search the datasets (opens new window); b) detect microbial peptides and determine the taxonomy associated with the peptides using Unipept; and c) validation of peptide spectral matches by using PepQuery (opens new window) and determining the number of valid peptides corresponding to microbial taxonomic units.The analysis of the respiratory tract samples using COMPIL 2.0 and Galaxy workflow (opens new window) with SearchGUI/PeptideShaker, Unipept and PepQuery resulted in detection of a few opportunistic pathogens (see table below).

# Workflow

The Galaxy workflow includes software tools to convert the input RAW files to MGF format. The MGF files are layer searched against the combined database of Human Uniprot proteome, UniProt database of clinically significant genera/species along with contaminant proteins and SARS-Cov-2 proteins database using X!tandem, MSGF+, OMSSA search algorithms (within SearchGUI) and FDR and protein grouping using PeptideShaker. The detected peptides were searched with Unipept 4.3 to obtain the taxonomic and functional information. Taxonomically relevant peptides were later subjected to analysis by PepQuery and Lorikeet to ascertain the quality of peptide identification.

# Results

Clinical studies from COVID-19 patients have reported co-infecting bacteria in COVID-19 patients. Interestingly, the PepQuery analysis supports the detection of these microbial peptides. We have followed this up with Lorikeet analysis to ascertain the spectral evidence.

Taxonomic Unit Pool 18 Pool 34 Pool 38 Pool 47 Pool 51
Coronaviridae 1 9 5 11 6
Anaerococcus 0 2 1 2 3
Dietzia 1 1 0 2 2
Prevotella buccalis 1 1 0 2 1
Candida albicans 1 0 2 0 0
Johnsonella ignava 1 0 1 2 0
Trichosporon asahii 0 0 0 2 0

Apart from Coronavirus peptides, we also detected peptides for Anaerococcus, Dietzia, Prevotella buccalis, Candida albicans, Johnsonella ignava and Trichosporon asahii. Anaerococcus species (opens new window) are anaerobic cocci that have been isolated in skin, vagina, and nasal cavity . Recent report (opens new window) has also demonstrated the isolation of Anaerococcus from bloodstream infection from an elderly immunocompetent person and shown that Anaerococcus are resistant to antimicrobial treatment. Dietzia species are aerobic, Gram-positive actinomycetes (opens new window) and have been implicated as potential invasive human pathogen and have been isolated from specimens taken from patients with acute infections . Dietzia species was also isolated from an immunocompromised patient with chronic obstructive pulmonary disease (COPD), and the infection is presumably related to the use of catheters (opens new window). Prevotella buccalis is an anaerobic bacterium isolated from dental plaque (opens new window). It was detected as one of the species associated with asthma patients based on nasal microbiome analysis (opens new window). Candida albicans is one of the most common fungal infections acquired during hospitalization and is prevalent in critically ill or immunocompromised patients (opens new window). Patients develop oropharyngeal candidiasis (opens new window), which can have an effect on the absorption of medication. Johnsonella ignava is an anaerobic Gram-negative rod-shaped bacterium isolated from human gingival crevices. In a study comparing oral microbiota (opens new window) in tumor and non-tumor tissues of patients with oral squamous cell carcinoma (OSCC), Johnsonella ignava was shown to be associated with the tumor site. Trichosporon asahii is an opportunistic fungal pathogen whose infections have increased in recent years, resulting in high mortality (opens new window) due to invasive infections in immunocompromised patients. It can cause infections (opens new window) on the skin, hair and lungs (chronic pneumonia).

This along with the analysis of gargling samples dataset (opens new window) analysis demonstrates the use of COMPIL 2.0 and metaproteomics workflow to detect any cohabitating emerging pathogens in COVID-19 patients using mass spectrometry based metaproteomics analysis.