Adding gene names to DESeq result
1
0
Entering edit mode
Chelsea • 0
@8759ce1a
Last seen 4 weeks ago
New Zealand

I am using the DESeq pipeline to create a heatmap of RNA transcripts deferentially expressed from 3 treatments, I have created a heatmap with the geneIDs but I am struggling to find a way to add gene names or alternatively produce a list of gene names that are deferentially expressed. I have a GFF3 file and a cds file for the genome I am using. I have tried the GenomicFeatures package but I cant access the vignettes for this package. Is there an updated package?



hse <- makeTranscriptDbFromGFF( "Red5_v1.0.gff3.gz", format="gff" )
Error in makeTranscriptDbFromGFF("Red5_v1.0.gff3.gz", format = "gff") :
could not find function "makeTranscriptDbFromGFF"

DESeq2 • 158 views
0
Entering edit mode

Hi, if it doesn't have to be GenomicFeature package, then there are plenty of other options. I'm usually using biomaRt R package.

res<-results(dds)

library(biomaRt)
mart <- useMart("ensembl", dataset = "mmusculus_gene_ensembl")

head(listAttributes(mart)) # here you can see which attributes are available

ensemble2gene <- getBM(attributes=c("external_gene_name","ensembl_gene_id"),
filters = "ensembl_gene_id",
values = rownames(res),
mart = mart)

match(ensemble2gene$ensembl_gene_id, rownames(res)) res$ensembl_gene<-rownames(res)
res<-merge(as.data.frame(res), ensemble2gene, by.x="ensembl_gene", by.y="ensembl_gene_id", all.y=T)


In this case, rownames of my results were ensembl_gene_id, but you can of course adjust it to the values you're having. Then if you need you can replace the rownames with your external_gene_name and delete the additional column, that was created after merging.

Of course, it doesn't fit if your target genome is not in the package, and it's the reason you're creating custom database.

Regards

Artem

0
Entering edit mode
Basti ▴ 700
@7d45153c
Last seen 14 hours ago
France

See answer here : Could not find function "makeTranscriptDbFromGFF"