Network controllability enrichment analysis reveals that SARS-CoV-2 infection tends to target indispensable nodes of a directed human protein-protein interaction network
This article has been Reviewed by the following groups
Listed in
- Evaluated articles (ScreenIT)
Abstract
The COVID-19 disease has been a global threat caused by the new coronavirus species, SARS-CoV-2, since early 2020 with an urgent need for therapeutic interventions. In order to provide insight into human proteins targeted by SARS-CoV-2, here we study a directed human protein-protein interaction network (dhPPIN) based on our previous work on network controllability of virus targets. We previously showed that human proteins targeted by viruses tend to be those whose removal in a dhPPIN requires more control of the network dynamics, which were classified as indispensable nodes. In this study we introduce a more comprehensive rank-based enrichment analysis of our previous dhPPIN for SARS-CoV-2 infection and show that SARS-CoV-2 also tends to target indispensable nodes in the dhPPIN using multiple proteomics datasets, supporting validity and generality of controllability analysis of viral infection in humans. Also, we find differential controllability among SARS-CoV-2, SARS-CoV-1, and MERS-CoV from a comparative proteomics study. Moreover, we show functional significance of indispensable nodes by analyzing heterogeneous datasets from a genome-wide CRISPR screening study, a time-course phosphoproteomics study, and a genome-wide association study. Specifically, we identify SARS-CoV-2 ORF3A as most frequently interacting with indispensable proteins in the dhPPIN, which are enriched in TGF-beta signaling and tend to be sources nodes and interact with each other. Finally, we built an integrated network model of ORF3A-interacting indispensable proteins with multiple functional supports to provide hypotheses for experimental validation as well as therapeutic opportunities. Therefore, a sub-network of indispensable proteins targeted by SARS-CoV-2 could serve as a prioritized network of drug targets and a basis for further functional and mechanistic studies from a network controllability perspective.
Article activity feed
-
SciScore for 10.1101/2021.04.18.440358: (What is this?)
Please note, not all rigor criteria are appropriate for all manuscripts.
Table 1: Rigor
NIH rigor criteria are not applicable to paper type.Table 2: Resources
Experimental Models: Cell Lines Sentences Resources CRISPR knock-out loss-of-function screening study in SARS-CoV-2 infection in human cells: We use multiple sets of results from a genome-wide CRISPR study in a human alveolar basal epithelial carcinoma cell line, A549, with ACE2 expression (Daniloski et al., 2020). A549suggested: NoneSARS-CoV-2 infection phosphoproteomics data: Bouhaddou et al. published time-course phosphoproteomics data in SARS-CoV-2 infection in Vero E6 cells and performed Gene Ontology enrichment analysis for human orthologues (Bouhaddou et al., 2020). Vero E6suggested: NoneSoftware and Algorithms Sentences Resources SciScore for 10.1101/2021.04.18.440358: (What is this?)
Please note, not all rigor criteria are appropriate for all manuscripts.
Table 1: Rigor
NIH rigor criteria are not applicable to paper type.Table 2: Resources
Experimental Models: Cell Lines Sentences Resources CRISPR knock-out loss-of-function screening study in SARS-CoV-2 infection in human cells: We use multiple sets of results from a genome-wide CRISPR study in a human alveolar basal epithelial carcinoma cell line, A549, with ACE2 expression (Daniloski et al., 2020). A549suggested: NoneSARS-CoV-2 infection phosphoproteomics data: Bouhaddou et al. published time-course phosphoproteomics data in SARS-CoV-2 infection in Vero E6 cells and performed Gene Ontology enrichment analysis for human orthologues (Bouhaddou et al., 2020). Vero E6suggested: NoneSoftware and Algorithms Sentences Resources virus-human protein-protein interactions data: We use data from the IntAct database (Orchard et al., 2014), which contains 2,020 virus-human PPIs between 26 SARS-CoV-2 proteins and 1,341 human proteins compiled from 5 different studies, as of June 17, 2020. IntActsuggested: (IntAct, RRID:SCR_006944)Network visualization: Network visualization was done using Cytoscape (Shannon et al., 2003). Cytoscapesuggested: (Cytoscape, RRID:SCR_003032)Results from OddPub: We did not detect open data. We also did not detect open code. Researchers are encouraged to share open data when possible (see Nature blog).
Results from LimitationRecognizer: We detected the following sentences addressing limitations in the study:One limitation is a lack of our understanding as to why there are such differences in enrichment of indispensable nodes between the high and low MOI conditions (Figs. 3C and 3D). As another functional relevance of indispensable proteins, we also investigated the protein phosphorylation data. As shown in Fig 4A, there is a clear difference of enrichment between the infection and non-infection samples for phosphorylation-regulated biological processes. However, post-infection differences over time are unclear. We do not expect to observe significant enrichment of indispensable nodes in all systems. For example, we previously observed that GWAS hits were not enriched in indispensable nodes (Vinayagam et al., 2016). We also observed that another GWAS study of SARS-CoV-2 infection in humans by Regeneron Inc. (Kosmicki et al., 2021) does not give rise to GWAS hits that are enriched in indispensable nodes (data not shown). Although we showed that the 23 GWAS hits by another study (Taylor et al., 2020) are enriched in indispensable nodes (Fig. 4B), the number of hits is relatively small, suggesting that statistical robustness may not be guaranteed. With those caveats and limitations in individual analyses of 7 different large-scale studies in mind, we identified those indispensable proteins with multiple supports of both physical and functional interactions with SARS-CoV-2 proteins (Figs. 5A and Fig. S5). We consider them as high confident with independent and complementary data supp...
Results from TrialIdentifier: No clinical trial numbers were referenced.
Results from Barzooka: We did not find any issues relating to the usage of bar graphs.
Results from JetFighter: We did not find any issues relating to colormaps.
Results from rtransparent:- Thank you for including a conflict of interest statement. Authors are encouraged to include this statement when submitting to a journal.
- No funding statement was detected.
- No protocol registration statement was detected.
-
