GAGE: No significant output while it should have
1
0
Entering edit mode
@littlefishes20-13054
Last seen 4.2 years ago

Hello

I usually analyse data with DEG I've got, and it is my first time using GAGE to analyse my data.

I took the output of DESeq2 (log2FoldChange) as input, and made a list of keggID-Geneset since there is no appropriate dataset in package suitable for my data.

I didn't get any significant result (q-value<0.05). It seems so strange, since I can get a several significant KEGG pathway while using DEGs. I wonder if my codes are wrong.

GAGE • 926 views
ADD COMMENT
0
Entering edit mode
#The codes are as follows
library(gage)
data <- read.table("config.FOfc.txt")
data2 <- read.delim("config.kegggs.txt",header = FALSE,sep = "\t",row.names=1)
data3 <- as.list(as.data.frame(t(data2)))
for (i in names(data3)){data3[[i]] = as.character(data3[[i]])}
for (i in names(data3)){data3[[i]] = unlist(strsplit(as.character(data3[[i]]), ' '))}
FO.kegg.p <- gage(data, gsets=data3,ref = NULL, samp = NULL, same.dir=F,rank.test=TRUE)
write.table(FO.kegg.p$greater,file="result-FO.kegg.p.greater.xls",sep="\t")
q()
ADD REPLY
0
Entering edit mode

#The input are shown below
> head(data)
                            lfc
HGL_N10009473        0.22861276
HGL_H00000302606-6   0.23119858
HGL_H00000368591     0.04589421
HGL_H00000355443-1           NA
HGL_H00000294923    -0.19674764
HGL_H00000326301-12          NA

>head(data3)

$`ko04724 Glutamatergic synapse`
  [1] "HGL_H00000312262"   "HGL_H00000263088"   "HGL_H00000342793"
  [4] "HGL_H00000317379"   "HGL_H00000310447"   "HGL_H00000307900-6"
  [7] "HGL_H00000307900-3" "HGL_H00000284384"   "HGL_H00000353362"
 [10] "HGL_H00000359719"   "HGL_H00000309591"   "HGL_H00000360021-1"
 [13] "HGL_H00000360021-2" "HGL_H00000380878"   "HGL_H00000378306"
 [16] "HGL_H00000378323"   "HGL_H00000263025-2" "HGL_H00000381803"
 [19] "HGL_H00000262493-1" "HGL_H00000416870-1" "HGL_H00000377503-2"
 [22] "HGL_H00000351635"   "HGL_N10016182"      "HGL_H00000355557"
 [25] "HGL_H00000371594"   "HGL_H00000375783"   "HGL_H00000248564"
 [28] "HGL_H00000248150"   "HGL_H00000388777"   "HGL_H00000300406"
 [31] "HGL_H00000282753"   "HGL_H00000306138"   "HGL_H00000378492"

ADD REPLY
0
Entering edit mode
Luo Weijun ★ 1.5k
@luo-weijun-1783
Last seen 3 months ago
United States

A few questions:

you read the fold changes into the data object. Does data include all genes or just a selected list of differentially expressed genes. GAGE (or similar methods like GSEA) expects all genes included, otherwise, your result would not be good.

Is data a 2 column matrix (data.frame), 1 column matrix or a vector? Are the gene IDs (HGL_N10009473 etc) the row names of the matrix or the names of the vector, as required by gage()? gene IDs can’t be column 1 of the matrix.

The way you read in KEGG gene set, i.e. read.delim() followed by as.list(as.data.frame()) seems to be problematic. You will get all gene set in the list) of the same length as it is converted from the rows of a matrix (data2). It doesn’t makes snese, since different pathways (or gene sets) definitely have different length.

ADD COMMENT

Login before adding your answer.

Traffic: 357 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