Correct way to make multiple comparisons on DESeq2?
1
0
Entering edit mode
jbnrodriguez ▴ 10
@dca7e9dc
Last seen 2.6 years ago
United Kingdom

I have a project where I have done RNA-seq (paired-end sequencing on Illumina HiSeq) of a worm at different days of development i.e. Ages 0-12. For each age, I have sequenced 3 replicate specimens. I'm completely new to DESeq2 and I was wondering if what I did below is correct.

library(DESeq2)
count_file_names <- grep("counts",list.files("featureCounts"),value=T)
specimen_type <- c("Age0","Age2","Age4","Age6","Age8","Age10","Age12")
sample_information <- data.frame(sampleName = count_file_names, fileName = count_file_names, condition = specimen_type)

Note- I switched to featureCounts instead of HTSeq for generating the counts but I formatted the featureCounts output files to reflect the HTSeq count output format. As far as I know, DESeq2 needs just counts and doesn't discriminate between the two count generating tools

DESeq_data <- DESeqDataSetFromHTSeqCount(sampleTable = sample_information, directory = "featureCounts", design = ~condition)
colData(DESeq_data)$condition <- factor(colData(DESeq_data)$condition,levels = c('Age0','Age2','Age4','Age6','Age8','Age10','Age12'))
rld <- rlogTransformation(DESeq_data, blind=T)

Note - I did relevel below based on Age0 even though it doesn't represent my control samples or anything. I wonder if this relevel step can be skipped though?

DESeq_data$condition <- relevel(DESeq_data$condition, "Age0")
DESeq_data <- DESeq(DESeq_data)

Below, I hope to know what genes are up-/down-regulated in Day12 worm specimens when compared to the rest. My real question is whether what I did below is correct/sensible? I plan to do in the same way for the other age specimens too.

Age_12_result <- results(DESeq_data, contrast = c(-1/7,-1/7,-1/7,-1/7,-1/7,-1/7,1/7), pAdjustMethod="BH", lfcThreshold=0, independentFiltering=T)
sig_de_results <- subset(Age_12_result, abs(log2FoldChange)> 1 & padj < 0.05)
sig_de_results <- sig_de_results[order(sig_de_results$padj, decreasing=F),]

My hope is that 'sig_de_results' tells me what genes were up-/down-regulated in Day12 specimens versus the average expression of genes in all of the remainder age specimens. Thanks a zillion in advance!

DESeq DESeq2 R • 2.2k views
ADD COMMENT
2
Entering edit mode
jbnrodriguez ▴ 10
@dca7e9dc
Last seen 2.6 years ago
United Kingdom

I solved this myself with the below code.

Compare age 12 versus age 0 with respect to the average of the rest

count_file_names <- grep("counts",list.files("featureCounts"),value=T)
specimen_type <- c("Age0","Age0","Age0","Age0","Age0","Age10","Age10","Age10","Age10","Age10","Age12","Age12","Age12","Age12","Age12","Age2","Age2","Age2","Age2","Age2","Age4","Age4","Age4","Age4","Age4","Age6","Age6","Age6","Age6","Age6","Age8","Age8","Age8","Age8","Age8")
sample_information <-data.frame(sampleName = count_file_names, fileName = count_file_names, condition = specimen_type)
DESeq_data <- DESeqDataSetFromHTSeqCount(sampleTable = sample_information, directory = "featureCounts", design = ~condition)
dds_lrt <- DESeq(DESeq_data, test="LRT", reduced = ~ 1)

To compare one condition versus the average of the rest

res12v0All_LRT <-  results(dds_lrt, contrast= list (c("condition_Age12_vs_Age0"), c("condition_Age10_vs_Age0","condition_Age2_vs_Age0","condition_Age4_vs_Age0","condition_Age6_vs_Age0","condition_Age8_vs_Age0")), listValues=c(1,-1/7))

When I do head(res12v0All_LRT), I get the below

log2 fold change (MLE): condition_Age12_vs_Age0 vs 0.143 condition_Age10_vs_Age0+condition_Age2_vs_Age0+condition_Age4_vs_Age0+condition_Age6_vs_Age0+condition_Age8_vs_Age0

LRT p-value: '~ condition' vs '~ 1'

DataFrame with 6 rows and 6 columns

sig_de_results <- subset(res12v0All_LRT, abs(log2FoldChange)> 1 & padj < 0.05)
sig_de_results <- sig_de_results[order(sig_de_results$padj, decreasing=F),]
ADD COMMENT
0
Entering edit mode

Note- I had more than 3 replicates in most of my sample age groups (I had stated the replicate number as being 3 in my initial question)

ADD REPLY

Login before adding your answer.

Traffic: 941 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