Brief Re-analysis of SARS-CoV-2 Proteomics from Bojkova et. al

Published on April 3rd, 2020
Written by Axel Martinelli
⏱ 5 min read

SARS proteomics data reanalysis plots with Omics Playground

Introduction

The first proteome of a SARS-CoV-2 in vitro infection has recently been published. The data set consists of a time series between 2h-24h post-infection. It is now available in our open access Viromics platform (now Omics Playground) for analysis. In their article, the authors describe several cellular pathways that are differentially expressed following infection and proceed to test potential antivirals based on the infection proteome profile.

We used our platform to re-analyse the data. To fully replicate the authors’ results we also included viral proteins in the analysis.

Clustering Analysis: exploring patterns in the data

First of all we generated a PCA plot, which showed that only the 24h infected sample was clearly clustering apart. Like the translatome PCA plot of the paper, we could also observe separate clusters for the infected samples at 6 h and 10h (Figure 1A).

The heatmap produced by the homonymous panel, indicates that protein expression changes only at 24h after infection. In particular, two clusters (named S1 and S2) show opposite patterns, with S2 being significantly downregulated and S1 significantly upregulated after 24h of infection compared to the other groups (Figure 1B).

Clustering Analysis. A) PCA plot of the study samples. Circled in yellow are the samples 24h post-infection, in green the infected samples after 10h, in grey infected samples after 6h and in purple and orange the rest. B) Clustered heatmap of the samples.
Figure 1. Clustering Analysis. A) PCA plot of the study samples. Circled in yellow are the samples 24h post-infection, in green the infected samples after 10h, in grey infected samples after 6h and in purple and orange the rest. B) Clustered heatmap of the samples.

The same pattern can also be observed with the plot obtained from the “Parallel” panel, which shows diverging trends for S1 and S2 proteins when comparing control and uninfected samples after 24h, but no major visual differences between control and infected samples at preceding time points (Figure 2).

Parallel Plot of the clusters of differentially expressed genes. Major trend differences in the S1 (Green) and S2 clusters (Orange) can be seen between infected and control groups at 24h post-infection.
Figure 2. Parallel Plot of the clusters of differentially expressed genes. Major trend differences in the S1 (Green) and S2 clusters (Orange) can be seen between infected and control groups at 24h post-infection.

The Functional annotation plot provides more information on the differentially expressed clusters. The downregulated cluster S2 included proteins related to adipogenesis and mitosis, whereas the upregulated S1 cluster included proteins related to TGF beta signalling, which is consistent with the main findings of the published article (Figure 3).

Functional annotation of the main clusters in the heatmap. Cluster S1 and S2 are associated with differential gene expression in the infected samples 24h post-inoculation.
Figure 3. Functional annotation of the main clusters in the heatmap. Cluster S1 and S2 are associated with differential gene expression in the infected samples 24h post-inoculation.

Screening for potential antiviral drugs

We next focused on screening for potential antiviral drugs based on the infected cells gene expression profiles by using the Drug Connectivity module. 

Using the 24h group for the analysis, the main mode of actions of potential inhibitors of the infection profiles included p38 MAPK inhibitors,src inhibitors, HDAC inhibitors and EGFR inhibitors (Figure 4), consistent with the results from previous posts (here and here) on other coronavirus species.

Figure 4. Mechanisms of actions of the drugs with potential inhibitory profiles against SARS-CoV-2 infections after 24h.
Figure 4. Mechanisms of actions of the drugs with potential inhibitory profiles against SARS-CoV-2 infections after 24h.

In terms of notable individual drugs, SB-202190, a p38-MAPK inhibitor with known antiviral properties (here, here and here) stood out. Fostamatinib, a Spleen Tyrosine Kinase (SYK) inhibitor was also highlighted. Intriguingly, SYK has been proposed as a target for treatment of inflammation in lung diseases, which fits the SARS-CoV-2 pathology.

However, no inhibitory potential was evident from the expression profile for trametinib, which we had previously suggested as effective against other coronavirus infections. This could be due to the different nature of the dataset (proteome) of SARS-CoV-2 compared to the previous studies (transcriptome).

You can read more about the role of trametinib in our previous post about finding potential biomarkers and therapeutic drugs with coronavirus datasets.

Scientist with lens exploring the drug connectivity analysis tab in Omics Playground

Unlock the full potential of your RNA-seq and proteomics data.

Try Omics Playground

About the Author

Axel Martinelli

Axel Martinelli’s academic background is in molecular biology and parasitology. He earned a Ph.D. on the genetics of strain-specific immunity against malaria infections and a master’s degree in bioinformatics with specialization in the analysis of omics data. During his postdoctoral career, he worked on genomics and transcriptomics studies and is currently the head of biology at Bigomics Analytics.