Similar protocols

To access compelling stats and trends, optimize your time and resources and pinpoint new correlations, you will need to subscribe to our premium service.


Pipeline publication

[…] ssues or cell type were discarded. Also, the datasets, which did not have raw data files, were discarded from the downstream analysis. Two datasets, one for pemphigus (GSE53873) and one for SLE (GDS4185), were included in this study. The covariate information available for the patients is summarized in Table S1 in Supplementary Material. Altogether 46 samples (4 PV, 15 PF, 13 SLE, and 14 healthy controls) were used in the analysis., To avoid a potential bias that could be introduced by obtaining two separate microarray datasets, the deposited gene expression data were directly used for batch normalization. The expression profiles were log2 transformed and batch normalization was done using “sva” and “combat” functions in SVA R package (). The effect of normalization was investigated by principal component analysis using the R-based “prcomp” function. Since batch normalization still produced biased results (Figure ), the raw files were preprocessed again and an additional normalization step was performed. In detail, raw gene expression profiles were deduced from text files (Codelink array) using Codelink R package (). Using the same package, first, the background was corrected with the “normexp” method and then normalized by the “cyclicloess” method. For Affymetrix data, raw gene expression for each sample was derived using R Affy package (). The background correction was performed by “backgroundCorrect (method = ‘normexp’)” and cyclic normalization was performed on log2 expression values using limma R package (). All the probes from each of the microarray platforms were filtered out for significant low expression/variation (P < 0.05) using the “varianceBasedfilter” function from DCGL R package (). The remaining probes were mapped to Ensembl gene identifiers and probes’ expression was collapsed to gene-level expression using “collapseRows” function with default parameters in WGCNA R package (). Consequently, batch normalization and statistical analysis were performed on the overlapping genes between two platfo […]

Pipeline specifications

Software tools sva, Codelink, affy