The Application of Consensus Weighted Gene Co-expression Network Analysis to Comparative Transcriptome Meta-Datasets of Multiple Sclerosis in Gray and White Matter

Abstract

Multiple sclerosis (MS) is a chronic inflammatory disease of the central nervous system characterized by demyelination, which leads to the formation of white matter lesions (WMLs) and gray matter lesions (GMLs). Recently, a large amount of transcriptomics or proteomics research works explored MS, but few studies focused on the differences and similarities between GMLs and WMLs in transcriptomics. Furthermore, there are astonishing pathological differences between WMLs and GMLs, for example, there are differences in the type and abundance of infiltrating immune cells between WMLs and GMLs. Here, we used consensus weighted gene co-expression network analysis (WGCNA), single-sample gene set enrichment analysis (ssGSEA), and machine learning methods to identify the transcriptomic differences and similarities of the MS between GMLs and WMLs, and to find the co-expression modules with significant differences or similarities between them. Through weighted co-expression network analysis and ssGSEA analysis, CD56 bright natural killer cell was identified as the key immune infiltration factor in MS, whether in GM or WM. We also found that the co-expression networks between the two groups are quite similar (density = 0.79), and 28 differentially expressed genes (DEGs) are distributed in the midnightblue module, which is most related to CD56 bright natural killer cell in GM. Simultaneously, we also found that there are huge disparities between the modules, such as divergences between darkred module and lightyellow module, and these divergences may be relevant to the functions of the genes in the modules.

Integrated Proteomics Identifies Neuritin (NRN1) as a Mediator of Cognitive Resilience to Alzheimer’s Disease

“…A correlation network was constructed using the consensus weighted gene co-expression network analysis (cWGCNA) algorithm, a systems biology approach to identify biologically meaningful, co-expression patterns (24, 25). The consensus configuration allows the identification of highly preserved modules, or clusters of interconnected proteins, shared across BA6 and BA37 while retaining region-specific relationships ( Fig.…”

“…We used the consensus Weighted Gene Correlation Network Analysis (cWGCNA; version 1.69) algorithm to generate a central network of co-expression modules from both brain regions (24, 25). The WGCNA::blockwiseConsensusModules function was run with soft threshold power at 7.0, deepsplit of 4, minimum module size of 30, merge cut height at 0.07, mean topological overlap matrix (TOM) denominator, using bicor correlation, signed network type, pamStage and pamRespectsDendro parameters both set to TRUE and a reassignment threshold of 0.05.…”

Brain proteome‐wide association study linking‐genes in multiple sclerosis pathogenesis

“…Several studies have explored differences of genetic effects between white and gray matter in MS and have reported the differences at both the proteome 13 and transcriptome levels. 58 , 59 This may be due to the various pathophysiological mechanisms of demyelination in gray and white matter, 60 as white matter lesions are accompanied by activation of local glial cells and infiltration of peripheral leucocytes, whereas gray matter lesions show a lack of activated glial cells and few infiltration of peripheral leucocytes. 61 Another finding was that risk gene dysregulation was associated with the degree of pathological changes, as up‐regulation of SHMT1 could only be observed in macroscopic lesions comparing with healthy controls, but not in NAWM.…”

“…Our study found that risk gene dysregulation had specificity in distribution, as SHMT1 and FAM120B were significantly up‐regulated in the white matter, while ICA1L was considerably down‐regulated in the gray matter. Several studies have explored differences of genetic effects between white and gray matter in MS and have reported the differences at both the proteome 13 and transcriptome levels 58,59 . This may be due to the various pathophysiological mechanisms of demyelination in gray and white matter, 60 as white matter lesions are accompanied by activation of local glial cells and infiltration of peripheral leucocytes, whereas gray matter lesions show a lack of activated glial cells and few infiltration of peripheral leucocytes 61 .…”

Integrated Proteomics to Understand the Role of Neuritin (NRN1) as a Mediator of Cognitive Resilience to Alzheimer’s Disease

“…We used the cWGCNA (version 1.69) algorithm to generate a central network of co-expression modules from both brain regions ( 30 , 31 ). The WGCNA::blockwiseConsensusModules function was run with soft threshold power at 7.0, deepsplit of 4, minimum module size of 30, merge cut height at 0.07, mean topological overlap matrix (TOM) denominator, using bicor correlation, signed network type, pamStage and pamRespectsDendro parameters both set to TRUE and a reassignment threshold of 0.05.…”