goseq with non native organism
1
0
Entering edit mode
3.8 years ago
alyyha • 0

I´m using goseq package after DESeq2 analysis and I am totally new in R

When I would like to test the pwf of my DE genes, I get an error:

if (!requireNamespace("BiocManager", quietly = TRUE))
install.packages("BiocManager")

BiocManager::install("TxDb.Athaliana.BioMart.plantsmart28")

library(TxDb.Athaliana.BioMart.plantsmart28)
txdb<-TxDb.Athaliana.BioMart.plantsmart28
txsByGene=transcriptsBy(txdb,"gene")
lengthData=median(width(txsByGene))



assayed.genes <- rownames(res)
de.genes <- rownames(res)[ which(res$padj < 0.05)]
genes<-as.integer(assayed.genes%in%de.genes)
names(genes)<-row.names(res$padj)

genes
    0     1 
29357  3476 

pwf<-nullp(genes,txdb,assayed.genes,lengthData)

Error in nullp(genes, txdb, assayed.genes, lengthData) : 
bias.data vector must have the same length as DEgenes vector!

length(lengthData)=33602

length(genes)=32833

DEgenes vector in the script is genes and contains all Arabidopsis genes (32833). I did not filter for counts.

bias.data vector in the script is lengthData and contains median length values of all transcripts (33602).

Does someone knows how to solve this error? Thank you very much in advance!

rna-seq goseq arath • 1.3k views
ADD COMMENT
0
Entering edit mode
3.8 years ago
ATpoint 83k

You have 32833 genes and 33602 elements in the bias vector, see the problem? You must have one value in the bias bector and it must have the exact same order as the gene list.

Code suggestion:

assayed.genes <- rownames(res)
genes         <- as.integer(res$padj < 0.05)
biasdata      <- lengthData[match(assayed.genes, names(txsByGene))]
nP            <- nullp(DEgenes = genes, bias.data = biasdata)
ADD COMMENT
0
Entering edit mode

Yes I see this problem, thank you. I am just wondering how to fix it because I don´t know how to do it.

ADD REPLY
0
Entering edit mode

Just added a suggestion.

ADD REPLY
0
Entering edit mode

Thank you very much! I'll try

ADD REPLY
0
Entering edit mode

Also added a command for nullp. The DEgenes is a vector of same length as biasdata and it has a 1 if the gene is DE and a 0 if not.

ADD REPLY
0
Entering edit mode

It worked. Danke schon!!

ADD REPLY
0
Entering edit mode

Glad to help! If the answer solved the issue please accept it to provide closure (using the checkmark button).

ADD REPLY

Login before adding your answer.

Traffic: 1055 users visited in the last hour
Help About
FAQ
Access RSS
API
Stats

Use of this site constitutes acceptance of our User Agreement and Privacy Policy.

Powered by the version 2.3.6