risserlin
commented
4 years ago At the minimum I think that the GEM file requires a pvalue and fdr value. (ie. it needs Name, description, pval, FDR) but it looks like you are only giving it the pvalue. I know that in the case of gprofiler the fdr is the pvalue but I usually just duplicate the value and put the pvalue in both the 3rd and 4th column.
Thank you for making a great App. I have tried to established a GEM in Cytoscape out of a .txt. file I obtained from gprofiler 2 R package using this code.
`gostres0 <- gost(query = rownames(sc_C0_markers), organism = "mmusculus", ordered_query = TRUE, multi_query = FALSE, significant = TRUE, exclude_iea = TRUE, measure_underrepresentation = FALSE, evcodes = TRUE, user_threshold = 0.05, correction_method = "g_SCS", domain_scope = "annotated", custom_bg = NULL, numeric_ns = "", sources = "GO", as_short_link = FALSE) sc_C0gprofiler_results <- gostres0$result
Order the results by p-adjusted value sc_C0gprofiler_results <- sc_C0gprofiler_results[order(sc_C0gprofiler_results$p_value), ] gostplot(gostres0, capped = TRUE, interactive = TRUE) p0 <- gostplot(gostres0, capped = TRUE, interactive = FALSE) pdf("sc_C0__gplot.pdf", width = 11, height =11 )
publish_gostplot(p0, highlight_terms = sc_C0gprofiler_results$term_id[1:19], width = 20, height = 20, filename = NULL) dev.off() get the GEM file for cytoscape gem0 <- gostres0$result[,c("term_id", "term_name", "p_value", "intersection")] colnames(gem0) <- c("GO.ID", "Description", "p.Val", "Genes") gem0$FDR <- gem0$p.Val write.table(gem0, file = "c0gProfiler_gem.txt", sep = "\t", quote = F, row.names = F)`
when Building the map I get an error message which states "Building EnrichmentMap parsing generic result file For input string: "Lpar1, Ifitmal,Vtn................"
Do you have any idea what might be the problem? Thank you very much
Cytoscape : Version: 3.8.0 R version 3.6.3 (2020-02-29)