SARS coronavirus. Viral and host pathways analyis

The aim of this publication is to identify viral and host pathways involved in coronavirus infection.

For this purpose, we have aggregated at the protein/gene level the predicted activity scores of 1.7M molecules on SARS/MERS viral assays, on the basis of their molecular interactions with target proteins recorded in ChEMBL DB. Click on the links above to access to detailed information about such procedures.

We can use low level pathways from Reactome Pathways DB to trace a gene/pathways association map as below.

We can use this gene/pathways association map to generate, by aggregation of values at the pathway level, several indexes that act as descriptors of the pathway involvement in coronavirus infection.

For the scatter plot below we have defined:

• log10SumPositiveInteractions.- As the logged addition of all active interactions of predicted SARS active molecules with all the proteins belonging each pathway.
• aggregatedSarsScore.- As the average value of predicted activity against SARS viruses aggregated at the pathway level from the values assigned to each protein/gene.
• [[activeTargetsInPathway]] / [proteinsInPathway]]].- Number of proteins considered relevant for SARS coronavirus infection in a particular pathway vs the total number of protein of that pathway referred in the ChEMBL DB.
• combinedScore.- Just as linear combination of the number of evidences supporting a score (log10SumPositiveInteractions) and the predicted score from aggregation for a pathway (aggregatedSarsScore).

And we also can use the gene scores and their association to their respective pathways to perform pathways enrichment, thus getting a n statistical indicator of enrichment (enrichmentScore) and a pValue of confidence (FDR: false discovery rate).

In this particular case, we have used Guangchuang Yu methodology published in An R package for Reactome Pathway Analysis. Additional details about the procedure can be obtained accessing to  Guangchuang’s published page through the link above.

Now, we can see both, enrichment scores and aggregation scores together in a bar chart.

And now, we can incorporate gene and pathway names and scores to the gene/pathway association network graph.

It may be a nice view, although there is not much info we can get from this view nor conclusions to infer.

So, let’s do two things: First, select most significant associations selecting best combinations of their aggregation and enrichment scores. Next, subset the network in different areas on the basis of the closeness of their respective interactions in order to define different independent network modules to perform a deeper analysis.

Now, we can look at the different modules closer. Just click on the network graphs below to expand views and on the yellow buttons to see a more complete analysis.

Pathways associated to ribosomal proteins 

Pathways associated to proteasomal subunits

Ribosomal proteins and proteasomal subunits are closely connected by common pathways.   

GPCR-associated pathways

PI3K/AKT signaling

BCL-2 inflammasome

Notch signaling

Cell cycle and recycling

Mitosis and viral infection

Cell differentiation

Autophagy and its regulators

RAS/RAF signaling

Calcium channels signaling

Coagulation, viral replication, ion pumps and intracellular trafficking

See most relevant gene/pathways associations and scores