RNA extraction, sequencing and analysis
PBMCs were isolated from the blood of RA patients and healthy donors via centrifugation over Ficoll-Paque Plus (GE Healthcare, Sweden). RNA was extracted from PBMCs using Trizol: chloroform, precipitated in isopropanol. Sample quality was assessed using Nanodrop. Total RNA was enriched for mRNA using poly-A selection. In brief, mRNA was fragmented, reverse transcribed, adapted with sequencing primers and sample barcodes, size selected, and PCR enriched. Libraries were sequenced on the HiSeq 2000 platform. RNA sequence reads were aligned to the human reference genome (NCBI, hg19) using TopHat2 aligners. Normalised gene counts were calculated using HTSeq, and differential gene expression between samples was identified via edgeR. Gene expression (RPKM) values were calculated using Cufflinks v2.0.2. Gene Ontology (http://geneontology.org/) associations were determined using Gorilla. Network construction was based on the Ingenuity Pathway Analysis (Qiagen, CA, USA) experimental evidence database.