Additional file 1: of Microbiome dysbiosis is associated with disease duration and increased inflammatory gene expression in systemic sclerosis skin

Table S1. Gene expression data. Normalized RNA-seq data were classified using a support vector machine trained using a merged and curated dataset composed of samples from GSE9285, GSE32413, and GSE45485. To visual results, the probe ID list from Johnson et al. [4] was collapsed on gene ID. This gene list was compared against normalized FPKM values for all 36 RNA-seq samples, resulting in a total of 1010 overlapping genes. Intrinsic subset assignments for individual genes are not possible based on the nature of gene expression in SSc patients. Each patientâ s intrinsic subset assignment was determined based on the collective co-expression of all 1010 genes in this dataset, with both high and low expression of individual genes important for determining subset distinctions. Furthermore, both high and low expression of individual genes often extends across multiple intrinsic subsets. This inherently prevents providing subset-specific calls for individual genes. (XLSX 418 kb)