SEQanswers

Go Back   SEQanswers > Applications Forums > RNA Sequencing



Similar Threads
Thread Thread Starter Forum Replies Last Post
DEG for paired samples, biological replicates hypatia RNA Sequencing 24 05-07-2014 02:25 AM
DEG analysis without gff/gtf file syintel87 Bioinformatics 6 01-11-2013 02:06 AM
advice on DESeq and EdgeR output from bacterial RNA-seq ugolino Bioinformatics 2 11-16-2012 12:21 PM
Doing DEG for a suspected new transcript? SamCurt Bioinformatics 0 12-29-2011 02:08 PM

Reply
 
Thread Tools
Old 10-21-2013, 02:26 AM   #1
bernardo_bello
Member
 
Location: Spain

Join Date: May 2012
Posts: 51
Smile Too many DEG with edgeR output?

Hi there,

I'm dealing with bacterial RNA-seq analysis. My experiment is very simple. Two samples to compare and no replicates. Reads were generated with Ion Torrent PGM using 316 chip. One for each sample and performed in different days.

Since I had too many differentially expressed genes, ┐Should I be more conservative assigning edgeR dispersion value? Also, there are considerable more up-regulated genes in *exvivo* than in plate sample.

See logFC_vs_logCPM figure:

https://docs.google.com/file/d/0B8-Z...it?usp=sharing

Thanks for you help, Bernardo



- tmap code:

tmap map2 -f HPNK_clean.fsa -r exvivo.fastq -i fastq -s exvivo.sam --verbose

tmap map2 -f HPNK_clean.fsa -r plate.fastq -i fastq -s plate.sam --verbose

- Flasgstat:

exvivo:

>3240242 + 0 in total (QC-passed reads + QC-failed reads)


>2132481 + 0 mapped (65.81%:nan%)

plate:

>3774075 + 0 in total (QC-passed reads + QC-failed reads)


>3510438 + 0 mapped (93.01%:nan%)


- count:

python -m HTSeq.scripts.count -m intersection-nonempty -t CDS -i ID exvivo.sam HPNK.gff > exvivo.counts

python -m HTSeq.scripts.count -m intersection-nonempty -t CDS -i ID plate.sam HPNK.gff > plate.counts


- count stats:

ex-vivo stats

>no_feature 777946

>ambiguous 1

>too_low_aQual 0

>not_aligned 1107761

>alignment_not_unique 0

plate stats

>no_feature 776707

>ambiguous 47

>too_low_aQual 0

>not_aligned 263637

>alignment_not_unique 0

- edgeR code:

library(edgeR)

files <- dir(pattern="*\\.counts$")

RG <- readDGE(files, header=FALSE)

RG

keep <- rowSums(cpm(RG)>1) >= 2 #we keep genes that achieve at least one count per million (cpm) in at least three samples

RG <- RG[keep,]

dim(RG)


RG <- calcNormFactors(RG)

RG$samples


plotMDS(RG)


bcv <- 0.2 #Assigned dispersion value of 0.2

m <- as.matrix(RG)

d <- DGEList(counts=m, group=(1:2)) #modify 'group' depending on sample number. Also can be adapted to replicated samples, see'?DGEList'.

d

et <- exactTest(d, pair=(1:2),dispersion=bcv^2) #exactTest(RG, pair=(1:2),dispersion=bcv^2)

et

top <- topTags(et)

top

cpm(RG)[rownames(top), ] #Check the individual cpm values for the top genes:

summary(de <- decideTestsDGE(et)) #The total number of DE genes at 5% FDR is given by'decideTestsDGE'.


[,1]

-1 200

0 1176

1 769

Of the 'number' tags identified as DE, 769 are up-regulated ex-vivo and 200 are down-regulated.

detags <- rownames(RG)[as.logical(de)] #detags contains the DE genes at 5% FDR

plotSmear(et, de.tags=detags) #plot all genes and highlight DE genes at 5% FDR

abline(h=c(-1, 1), col="blue") #The blue lines indicate 2-fold changes.

title("plate vs ex-vivo")

dev.copy2pdf(file = "Figure_1.pdf") #Save as .pdf##
bernardo_bello is offline   Reply With Quote
Reply

Tags
bacteria, edger, iontorrent, rna-seq, tmap

Thread Tools

Posting Rules
You may not post new threads
You may not post replies
You may not post attachments
You may not edit your posts

BB code is On
Smilies are On
[IMG] code is On
HTML code is Off




All times are GMT -8. The time now is 03:08 AM.


Powered by vBulletin® Version 3.8.9
Copyright ©2000 - 2018, vBulletin Solutions, Inc.
Single Sign On provided by vBSSO