diff --git a/R/cite_RNAseqGSEA.R b/R/cite_RNAseqGSEA.R
index ee79528..affb5af 100644
--- a/R/cite_RNAseqGSEA.R
+++ b/R/cite_RNAseqGSEA.R
@@ -30,7 +30,7 @@ cite_RNAseqGSEA <- function(fold_changes = 1.5, normalisation_type = "LRT_ruvR_u
 
     cat(stringr::str_interp('Methods: RNA seq and GSEA processing
 
-RNAseq data was trimmed using cutadapt[${getIndex(citations, "cutadapt")}] v1.18 and fastQC[${getIndex(citations, "fastqc")}] v0.11.9. Mapping was done with Homo_sapiens.GRCh38.101.gtf[${getIndex(citations, "refgenome")}] as a reference genome. Trim and mapping quality was assesed with the multiqc[${getIndex(citations, "multiqc")}] utility version 1.8. Differential expression analysis was done with use of the edgeR[${getIndex(citations, "edger")}] package version 3.32.1 and EDAseq[${getIndex(citations, "edaseq")}] 2.24.0. An FDR cutoff of 0.05 was selected and fold change cutoff: ${if(length(fold_changes) == 1) fold_changes else paste0(fold_changes, ", ")}; ${if(normalisation_type == "LRT_ruvR_upperquartile") "LRT (likelihood ratio test) RUVr (remove unwanted variation) upperquartile normalisation was used" else if(normalisation_type == "TMM") "TMM normalisation was used"}. GSEA[${getIndex(citations, "gsea")}, ${getIndex(citations, "msigdb")}] (gene set enrichment analysis) was run with GSEA version 3.0. We used msigdb[${getIndex(citations, "msigdb")}, ${getIndex(citations, "msigdb")}] 7.3 human gene set files including: c2.cp.kegg.v7.3.symbols.gmt, c2.cp.reactome.v7.3.symbols.gmt, c5.go.bp.v7.3.symbols.gmt, h.all.v7.3.symbols.gmt as reference pathways. Produced reports were filtered for an FDR cutoff of 0.25, these were then used to create heatmaps.
+RNAseq data was trimmed using cutadapt[${getIndex(citations, "cutadapt")}] v1.18 and fastQC[${getIndex(citations, "fastqc")}] v0.11.9. Mapping was done with Homo_sapiens.GRCh38.101.gtf[${getIndex(citations, "refgenome")}] as a reference genome. Trim and mapping quality was assessed with the multiqc[${getIndex(citations, "multiqc")}] utility version 1.8. Differential expression analysis was done with use of the edgeR[${getIndex(citations, "edger")}] package version 3.32.1 and EDAseq[${getIndex(citations, "edaseq")}] 2.24.0. An FDR cutoff of 0.05 was selected and fold change cutoff: ${if(length(fold_changes) == 1) fold_changes else paste0(fold_changes, ", ")}; ${if(normalisation_type == "LRT_ruvR_upperquartile") "LRT (likelihood ratio test) RUVr (remove unwanted variation) upperquartile normalization was used" else if(normalisation_type == "TMM") "TMM normalization was used"}. GSEA[${getIndex(citations, "gsea")}, ${getIndex(citations, "msigdb")}] (gene set enrichment analysis) was run with GSEA version 3.0. We used msigdb[${getIndex(citations, "msigdb")}, ${getIndex(citations, "msigdb")}] 7.3 human gene set files including: c2.cp.kegg.v7.3.symbols.gmt, c2.cp.reactome.v7.3.symbols.gmt, c5.go.bp.v7.3.symbols.gmt, h.all.v7.3.symbols.gmt as reference pathways. Produced reports were filtered for an FDR cutoff of 0.25, these were then used to create heatmaps.
 
 ${paste0(paste(paste0(paste0("[", 1:length(citations)), "]"), citations, sep = " "), collapse = "\n")}'))
 }