AUTHOR=Chen Hong , Chen Chunqiu , Yuan Xiaoqi , Xu Weiwei , Yang Mu-qing , Li Qiwei , Shen Zhenyu , Yin Lu 

TITLE=Identification of Immune Cell Landscape and Construction of a Novel Diagnostic Nomogram for Crohn’s Disease

JOURNAL=Frontiers in Genetics

VOLUME=Volume 11 - 2020

YEAR=2020

URL=https://www.frontiersin.org/journals/genetics/articles/10.3389/fgene.2020.00423

DOI=10.3389/fgene.2020.00423

ISSN=1664-8021

ABSTRACT=Background: Crohn’s disease (CD) has an increasing incidence and prevalence worldwide. The etiology of CD is unclear and there is no gold standard for diagnosis. The dysregulated immune response and different infiltration status of immune cells are critical for CD pathogenesis. Thus, it is important to provide an overview of the immune cell alteration in CD and explore a novel method for auxiliary diagnosis.
Method: Microarray datasets (GSE112366 and GSE75214) were downloaded from Gene Expression Omnibus (GEO) and integrated for further analysis. An extended version of Cell-type Identification By Estimating Relative Subsets Of RNA Transcripts (CIBERSORTx) was utilized to estimate the fraction of 22 kinds of immune cells. Differentially expressed genes (DEGs) and a protein-protein interaction (PPI) network were identified. Gene set enrichment analysis (GSEA) and gene set variation analysis (GSVA) were performed to identify the differentially regulated pathways in CD. Co-expression patterns were constructed based on the correlation between immune cells, hub genes, and signaling pathways. Finally, the least absolute shrinkage and selection operator (LASSO) regression was conducted to verify the following logistic regression was not overfitting and a diagnostic nomogram for CD was built.
Results: The proportion of 17 immune cell types were found significantly altered between CD and normal control. A total of 150 differentially expressed genes were identified, which were mostly related to immune response. Among the 15 hub genes based on the PPI network, CXCL8 and IL1B had the highest interaction degree. GSEA and GSVA showed five pathways with significantly enriched status, of which the NOD-like receptor signaling pathway was crucial in the development of CD. Six variables comprise of CXCL8, IL1B, macrophage M1, Treg, CD8+ T cell and plasma cell were identified by LASSO regression and incorporated into a logistic regression model. The nomogram displayed a good prediction with an 0.915 under area of receiver operating curve (ROC). The C-index was 0.915 (95%CI: 0.875-0.955), which suggested good discrimination of the model.
Conclusion: Our results provide novel insight into cellular and molecular characteristics of CD in silico. These might offer potential biomarkers for diagnosis and targeted therapy.