add analysis of differences in expression assay and protocol

assays/AnalysisOfDifferencesInExpression/protocols/AnalysingDifferencesInExpression.md

+## Analysis of differences in expression
+
+To identify the factors associated with differences in expression of *R*-genes across transcriptomes, we performed an ANOSIM in Primer 7.0.13 (PRIMER-e; Figure S2). ANOSIM is a non-parametric statistical test similar to ANOVA. The starting point of the analysis is a pairwise dissimilarity matrix. In our case, the dissimilarity matrix was computed as follows: First the TPM values for each gene within each transcriptome were LOG (x+1) transformed. On the basis of these transformed TPM values, the dissimilarity in gene expression patterns between transcriptomes were calculated based on Euclidean distances. Ranking was applied to the distance matrix. The two libraries from potato (SRR6511453 and ERR791944) with exceptionally low expression of the entire *R*-gene repertoire were excluded in these analyses.
+
+To determine if gene expression is more similar within groups than between groups (for example when groups are defined by infection status or tissue type) the R test statistic value was calculated. The R values can range from -1 to 1, with larger values corresponding to greater differences between groups. Statistical significance is calculated through permutation of the group labels and recalculation of the R value for each replicate. In our case, 999 permutations were generated. The following factors were evaluated: BioProject, tissue type, type of treatment, specific treatment organism, life cycle of the organism, type/kingdom of the organism, susceptible vs. resistant cultivar, relative sequencing depth, paired- or single-end sequencing and days post infection. The ANOSIM analysis was also applied to the differential gene expression data (see below).
\ No newline at end of file