CoChat

Understanding the molecular drivers of disease heterogeneity in Crohn's disease using multi-omic data integration and network analysis

Open access

The DOI matches its Crossref record; 105 of the 105 references it cites were found in a registry or an index.

Abstract

Crohn’s disease (CD), a form of inflammatory bowel disease (IBD), is characterized by heterogeneity along multiple clinical axes, which in turn impacts disease progression and treatment modalities. Using advanced data integration approaches and systems biology tools, we studied the contribution of CD susceptibility variants and gene expression in distinct peripheral immune cell subsets (CD14+ monocytes and CD4+ T cells) to relevant clinical traits. Our analyses revealed that most clinical traits capturing CD heterogeneity could be associated with CD14+ and CD4+ gene expression rather than disease susceptibility variants. By disentangling the sources of variation, we identified molecular features that could potentially be driving the heterogeneity of various clinical traits of CD patients. Further downstream analyses identified contextual hub proteins such as genes encoding barrier functions, antimicrobial peptides, chemokines, and their receptors, which are either targeted by drugs used in CD or other inflammatory diseases or are relevant to the biological functions implicated in disease pathology. These hubs could be used as cell type–specific targets to treat specific subtypes of CD patients in a more individualized approach based on the underlying biology driving their disease subtypes. Our study highlights the importance of data integration and systems approaches to investigate complex and heterogeneous diseases such as IBD.

Overview

The claims and statistics of this paper haven’t been checked yet.

Free: it reads the open-access text and checks the first 5 claims that cite a source we can check. Create a free account to check every claim.

A few free checks a day · papers already verified open instantly · results are public

Integrity signals

Registry

Crossref · 10.1093/ibd/izaa281

Cites

105 of 105 references verified0 citation errors · 0 not checkable

Statistics

Figures

FIGURE 1. A, Illustrative representation of the workflow used in the study to analyze and integrate the data sets. B, Variance inferred per latent factor and distributed across the data sets as inferred by Multi-Omic Factor Analysis.16
FIGURE 1. A, Illustrative representation of the workflow used in the study to analyze and integrate the data sets. B, Variance inferred per latent factor and distributed across the data sets as inferred by Multi-Omic Factor Analysis.16CC BY-NC · PubMed Central
FIGURE 2. A, Table depicting the explanatory LFs, which are defined as LFs associated with at least one clinical trait. The figure also shows the respective data sets that contribute to the variance explained by the LFs. By tracing the variance contributions to the LFs associated with the clinical traits, we could identify the -omic data sets contributing to the clinical trait. Latent factors associated with gender were discarded to exclude sex-specific signals. B, Summary of modules (derived from co-expression analysis, see methods section “Bioinformatic analyses” for more details), their relevance to the corresponding phenotypes (indicated by the column “Module activity in phenotype”), and the overlap between the hubs (top 50 genes) in each of the modules and targets of drugs used in intestinal inflammatory disorders such as Crohn’s disease and ulcerative colitis, other inflammatory disorders such as ankylosing spondylitis, rheumatoid arthritis, psoriasis, and sclerosing cholangitis, associations with CD, role as CD eQTLs or genes relevant to CD in terms of their activity (antimicrobial peptide functions, barrier functions, chemokine functions, and cell adhesion). ^ - if the total number of genes in a module was less than 50, all the genes in the module were considered as hubs. In the last column, the active modules with the highest percentage of unique overlapping hubs are indicated.
FIGURE 2. A, Table depicting the explanatory LFs, which are defined as LFs associated with at least one clinical trait. The figure also shows the respective data sets that contribute to the variance explained by the LFs. By tracing the variance contributions to the LFs associated with the clinical traits, we could identify the -omic data sets contributing to the clinical trait. Latent factors associated with gender were discarded to exclude sex-specific signals. B, Summary of modules (derived from co-expression analysis, see methods section “Bioinformatic analyses” for more details), their relevance to the corresponding phenotypes (indicated by the column “Module activity in phenotype”), and the overlap between the hubs (top 50 genes) in each of the modules and targets of drugs used in intestinal inflammatory disorders such as Crohn’s disease and ulcerative colitis, other inflammatory disorders such as ankylosing spondylitis, rheumatoid arthritis, psoriasis, and sclerosing cholangitis, associations with CD, role as CD eQTLs or genes relevant to CD in terms of their activity (antimicrobial peptide functions, barrier functions, chemokine functions, and cell adhesion). ^ - if the total number of genes in a module was less than 50, all the genes in the module were considered as hubs. In the last column, the active modules with the highest percentage of unique overlapping hubs are indicated.CC BY-NC · PubMed Central
FIGURE 3. A, The “scale-free” aspect of the modules (inferred from the monocyte gene expression) obtained over a window of values for the soft threshold beta parameter. Modules corresponding to the beta value of 8 was chosen because there was a plateauing of the observed “scale-free”-ness or aspect. Many real biological networks display scale-free behavior captured by the scale-free topology model fit R2 value plotted on the y axis. B, The activity profile of the monocyte gene expression modules in relation to the disease behavioral phenotype. Only the modules active with respect to the phenotype are displayed. The NES score represents the z-score normalized expression of the samples within each phenotype and is used to assess the alterations of module activity with respect to the phenotype. C, Overrepresented Reactome gene sets at FDR ≤ 0.05 in module 5, (D) in module 8 (E), and in module 9. F, The top 10 genes (hubs) identified in module 9 after integrating the co-expression network derived from module 9 with the benchmarked interaction network PCNet.43 Hubs are defined as nodes (genes) with a high connectivity (ie, number of neighbors) in the network and represent nodes that could play an active role in the given context.
FIGURE 3. A, The “scale-free” aspect of the modules (inferred from the monocyte gene expression) obtained over a window of values for the soft threshold beta parameter. Modules corresponding to the beta value of 8 was chosen because there was a plateauing of the observed “scale-free”-ness or aspect. Many real biological networks display scale-free behavior captured by the scale-free topology model fit R2 value plotted on the y axis. B, The activity profile of the monocyte gene expression modules in relation to the disease behavioral phenotype. Only the modules active with respect to the phenotype are displayed. The NES score represents the z-score normalized expression of the samples within each phenotype and is used to assess the alterations of module activity with respect to the phenotype. C, Overrepresented Reactome gene sets at FDR ≤ 0.05 in module 5, (D) in module 8 (E), and in module 9. F, The top 10 genes (hubs) identified in module 9 after integrating the co-expression network derived from module 9 with the benchmarked interaction network PCNet.43 Hubs are defined as nodes (genes) with a high connectivity (ie, number of neighbors) in the network and represent nodes that could play an active role in the given context.CC BY-NC · PubMed Central
FIGURE 4. A, The activity profile of the CD4+ gene expression modules in relation to the disease behavioral phenotype. Only the modules active with respect to the phenotype are displayed. B, Overrepresented Reactome gene sets at FDR ≤ 0.05 in module 2. C, The top 10 genes (hubs) identified in module 2 and (D) module 3 after integrating the co-expression network derived from the corresponding modules with the benchmarked interaction network PCNet.43 Genes encoding proteins such as OGT (UDP-N-acetylglucosamine--peptide N-acetylglucosaminyltransferase) and which play important roles in mediating T-cell responses83 are identified as hubs.
FIGURE 4. A, The activity profile of the CD4+ gene expression modules in relation to the disease behavioral phenotype. Only the modules active with respect to the phenotype are displayed. B, Overrepresented Reactome gene sets at FDR ≤ 0.05 in module 2. C, The top 10 genes (hubs) identified in module 2 and (D) module 3 after integrating the co-expression network derived from the corresponding modules with the benchmarked interaction network PCNet.43 Genes encoding proteins such as OGT (UDP-N-acetylglucosamine--peptide N-acetylglucosaminyltransferase) and which play important roles in mediating T-cell responses83 are identified as hubs.CC BY-NC · PubMed Central
FIGURE 5. The functional profile of the genes in the monocyte-gene expression derived module 9 active across the disease behavioral phenotype. Sixty-four percent of the module 9 hubs were annotated as being targets of drugs used to treat intestinal inflammatory disorders such as CD and UC, other inflammatory disorders such as ankylosing spondylitis, rheumatoid arthritis, psoriasis, and sclerosing cholangitis, associations with CD, role as CD eQTLs or genes relevant to CD in terms of their activity (antimicrobial peptide functions, barrier functions, chemokine functions and cell adhesion). Genes marked in orange denote transcription factors. **Indicates TFs that are identified as being relevant regulators (of the genes in the same modules) by ChEA3.37 The figure indicates that more than half of the genes were previously identified as involved in the pathogenesis of CD as recorded in the Open Targets database.48
FIGURE 5. The functional profile of the genes in the monocyte-gene expression derived module 9 active across the disease behavioral phenotype. Sixty-four percent of the module 9 hubs were annotated as being targets of drugs used to treat intestinal inflammatory disorders such as CD and UC, other inflammatory disorders such as ankylosing spondylitis, rheumatoid arthritis, psoriasis, and sclerosing cholangitis, associations with CD, role as CD eQTLs or genes relevant to CD in terms of their activity (antimicrobial peptide functions, barrier functions, chemokine functions and cell adhesion). Genes marked in orange denote transcription factors. **Indicates TFs that are identified as being relevant regulators (of the genes in the same modules) by ChEA3.37 The figure indicates that more than half of the genes were previously identified as involved in the pathogenesis of CD as recorded in the Open Targets database.48CC BY-NC · PubMed Central
FIGURE 6. A, Overlap of the transcription factors inferred for the active modules derived from gene expression in monocytes and CD4+ cells for the associated phenotypes. B, Transcriptional regulatory network showing the TFs modulating the expression of the monocyte modules active with respect to the disease behavior phenotype. Orange nodes represent the modules, purple nodes the TFs exclusively regulating the modules, and blue nodes the TFs regulating more than one module. The numbers within parentheses inside the orange nodes represent the number of Reactome signaling pathways enriched (FDR ≤ 0.05) within the genes found in the module. The thickness of the edges denote the percentage of genes of the module which are regulated by the TF. The figure indicates that the functionally distinct modules (as shown earlier by the set of Reactome pathways uniquely overrepresented in each of them) also have mostly distinct regulatory control, with very few transcription factors controlling the multiple modules. Visualization was performed using Cytoscape.90 C, Bar chart displaying the integrated mean ranks (from across orthogonal TF-target libraries) of the top 10 TFs for the genes in module 9 (enriched in interferon-gamma related pathways) derived from monocyte gene expression and active with respect to the disease behavioral phenotype. D, The distribution profile of genes in module 9 as targets of the most relevant TFs identified.
FIGURE 6. A, Overlap of the transcription factors inferred for the active modules derived from gene expression in monocytes and CD4+ cells for the associated phenotypes. B, Transcriptional regulatory network showing the TFs modulating the expression of the monocyte modules active with respect to the disease behavior phenotype. Orange nodes represent the modules, purple nodes the TFs exclusively regulating the modules, and blue nodes the TFs regulating more than one module. The numbers within parentheses inside the orange nodes represent the number of Reactome signaling pathways enriched (FDR ≤ 0.05) within the genes found in the module. The thickness of the edges denote the percentage of genes of the module which are regulated by the TF. The figure indicates that the functionally distinct modules (as shown earlier by the set of Reactome pathways uniquely overrepresented in each of them) also have mostly distinct regulatory control, with very few transcription factors controlling the multiple modules. Visualization was performed using Cytoscape.90 C, Bar chart displaying the integrated mean ranks (from across orthogonal TF-target libraries) of the top 10 TFs for the genes in module 9 (enriched in interferon-gamma related pathways) derived from monocyte gene expression and active with respect to the disease behavioral phenotype. D, The distribution profile of genes in module 9 as targets of the most relevant TFs identified.CC BY-NC · PubMed Central
FIGURE 7. Graphical summary of the clinical, experimental, and bioinformatic workflow used in the study. Some of the prominent results are also depicted.
FIGURE 7. Graphical summary of the clinical, experimental, and bioinformatic workflow used in the study. Some of the prominent results are also depicted.CC BY-NC · PubMed Central

Verification

Claims the paper makes

The claims and statistics of this paper haven’t been checked yet.

Free: it reads the open-access text and checks the first 5 claims that cite a source we can check. Create a free account to check every claim.

A few free checks a day · papers already verified open instantly · results are public

Claims other papers make about this one

Each is checked against this paper’s own text: does it say what the citing paper says it does?

Self-check

Percentages, sums and reported differences re-computed from the paper’s own numbers.

The self-check needs the paper’s full text.

Citations

Cites

Cited by

See all 53

Metadata

Authors
Sudhakar, P., Verstockt, B., Cremer, J., Verstockt, S., Sabino, J., Ferrante, M., & Vermeire, S.
Affiliations
  • Department of Chronic Diseases, Metabolism and Ageing, Translational Research Center for Gastrointestinal Disorders (TARGID)
  • Universitair Ziekenhuis Leuven (BE)
  • KU Leuven (BE)
Corresponding author
Vermeire, Séverine
Venue
Inflammatory Bowel Diseases
Published
2020-12-14
Volume
27
Issue
6
Pages
870-886
ISSN
  • 1078-0998 (ISSN-L)
  • 1536-4844
OpenAlex
W3112595545
Cited by
49
References
105
Citations by year
2026: 6 · 2025: 10 · 2024: 9 · 2023: 11 · 2022: 10 · 2021: 3
FWCI
3.51
Type
journal article
Language
en
Open access
Locations
  • Inflammatory Bowel Diseases (publishedVersion, cc-by-nc)
  • PubMed Central (publishedVersion, cc-by-nc)
  • PubMed (publishedVersion)
  • Lirias (KU Leuven) (acceptedVersion, cc-by-nc)
  • Lirias (KU Leuven) (submittedVersion, other-oa)
Funders
European Research Council (ERC-2015-AdG, 694679)
Topics
  • Health Sciences › Medicine › Gastroenterology › Inflammatory Bowel Disease
  • Life Sciences › Biochemistry, Genetics and Molecular Biology › Molecular Biology › Gut microbiota and health
  • Life Sciences › Biochemistry, Genetics and Molecular Biology › Genetics › Genetic Associations and Epidemiology
Keywords
Crohn's disease, disease heterogeneity
MeSH terms
Antimicrobial Peptides, Crohn Disease, Humans, CD4-Positive T-Lymphocytes, Chemokines, Systems Biology

About these checks

Registry, retraction, Cites and statistics results are computed, not judged, with one exception: a printed reference that no index lists is read once by a model before it counts as a citation error. Claims are extracted by a model and kept only when their quote appears verbatim in the text; each cited source is then checked, with a model’s help, for whether it says what the claim relies on. Those verdicts can be wrong — read the claim’s quote and the cited source before relying on one.

Statistics and arithmetic are recomputed, not judged.

Confidence bands are shown only for fields calibrated on our annotated test set. None is calibrated yet, so no band is shown.

What counts as a claim ›How we measure confidence ›How we check this ›

Are you an author, or do you see a problem with this page? Request a review

Data sources

Bibliographic data: Semantic Scholar, OpenAlex, Crossref.

Contains information from the Semantic Scholar Academic Graph Datasets, made available under the ODC Attribution License (ODC-BY 1.0).

Semantic Scholar’s data is described in: Kinney, R., et al. (2023). The Semantic Scholar Open Data Platform. arXiv:2301.10140.