Skip to Content

clusterProfiler introduction failed, 发布之日即是死亡之日

Posted on One min read

这篇旧博客写的本是一篇介绍 clusterProfiler 的文章失效的事，但我后来整理时回想起来我以前说过的一句话：

The day a research project is published shouldn’t be the day it’s dead.

Original post: use clusterProfiler as an universal enrichment analysis tool

But things allways change

  1. data link broke

    http://www.disgenet.org/ds/DisGeNET/results/all_gene_disease_associations.tar.gz should be replaced by http://www.disgenet.org/ds/DisGeNET/results/all_gene_disease_associations.tsv.gz

  2. use tibble::as_tibble() instead of DOSE::summary()

    Warning message:
In DOSE::summary(edn) :
  summary method to convert the object to data.frame is deprecated, please use as.data.frame instead.

  3. clusterProfiler::GSEA() failed

Following is updated code:

data("geneList", package = 'DOSE');
deg <- geneList %>% {names(.)[abs(.) > 2]};

agda <- readr::read_tsv('all_gene_disease_associations.tsv.gz');
disease2gene <- dplyr::select(agda, 'diseaseId', 'geneId');
disease2name <- dplyr::select(agda, 'diseaseId', 'diseaseName');

e <- clusterProfiler::enricher(deg, TERM2GENE = disease2gene, TERM2NAME = disease2name);
tibble::as_tibble(e);

DOSE:::barplot.enrichResult(e);
DOSE::dotplot(e);
DOSE::enrichMap(e);

g <- clusterProfiler::GSEA(geneList, TERM2GENE = disease2gene, TERM2NAME = disease2name);
tibble::as_tibble(g)