Register FAQ Members List Calendar Search Today's Posts Mark Forums Read

 Similar Threads Thread Thread Starter Forum Replies Last Post umnklang Bioinformatics 1 10-09-2012 06:48 PM Amative RNA Sequencing 2 01-16-2012 10:55 AM newbietonextgen Bioinformatics 1 12-30-2010 09:16 AM MerFer Bioinformatics 3 02-25-2010 01:48 AM mmanrique Bioinformatics 10 02-12-2010 03:13 PM

 11-20-2010, 05:26 PM #121 Sol Member   Location: Brazil Join Date: Oct 2010 Posts: 13 I would like know what the letters NA means as a result of DEGSeq. Another question: log2 is two fold change or four fold change thanks
11-20-2010, 08:03 PM   #122
Xi Wang
Senior Member

Location: MDC, Berlin, Germany

Join Date: Oct 2009
Posts: 317

Quote:
 Originally Posted by Sol I would like know what the letters NA means as a result of DEGSeq. Another question: log2 is two fold change or four fold change thanks

NA: when the read counts for a gene in both samples are zero, or zero and a small number (say, <5), the program will not calculate the values (such as fold-change, p-value) for this gene. "NA"s appear in those places.

log2 means base-2 logarithm. So
if fold-change = 1, log2(fold-change) = 0;
if fold-change = 2, log2(fold-change) = 1;
if fold-change = 4, log2(fold-change) = 2;
if fold-change = 0.5, log2(fold-change) = -1.
__________________
Xi Wang

 11-22-2010, 05:43 PM #123 Sol Member   Location: Brazil Join Date: Oct 2010 Posts: 13 How do you do to calculated the cutoff in the value the DEGseq, in pvalue. cutoff = 2 for example thanks
11-22-2010, 06:32 PM   #124
Xi Wang
Senior Member

Location: MDC, Berlin, Germany

Join Date: Oct 2009
Posts: 317

Quote:
 Originally Posted by Sol How do you do to calculated the cutoff in the value the DEGseq, in pvalue. cutoff = 2 for example thanks
The cufoffs are specified by users. If you ask how to calculate the p-values, please refer to our paper: http://bioinformatics.oxfordjournals.../full/26/1/136

BTW, p-value should be any real number between 0 and 1.
__________________
Xi Wang

 11-23-2010, 08:31 PM #125 wdt Member   Location: Southwest Join Date: Oct 2009 Posts: 19 HI, Using the sam to bed Perl script, I got the file like chr1 435837 435913 U0 0 + chr1 435837 435913 U0 0 - chr1 435837 435913 U1 0 - chr1 435838 435914 U1 0 + chr1 435838 435914 U1 0 - chr1 435838 435914 U1 0 - chr1 435840 435916 U2 0 - chr1 435840 435916 U2 0 - chr1 435840 435916 U3 0 - chr1 435840 435916 U2 0 - chr1 435842 435918 U4 0 - chr1 435842 435918 U4 0 - chr1 435844 435920 U2 0 - chr1 435844 435920 U2 0 - chr1 437189 437265 U2 0 + Could someone explain how U0, U1, U2 are assigned and what they are? Thanks,
11-23-2010, 08:54 PM   #126
Xi Wang
Senior Member

Location: MDC, Berlin, Germany

Join Date: Oct 2009
Posts: 317

Quote:
 Originally Posted by wdt HI, Using the sam to bed Perl script, I got the file like chr1 435837 435913 U0 0 + chr1 435837 435913 U0 0 - chr1 435837 435913 U1 0 - chr1 435838 435914 U1 0 + chr1 435838 435914 U1 0 - chr1 435838 435914 U1 0 - chr1 435840 435916 U2 0 - chr1 435840 435916 U2 0 - chr1 435840 435916 U3 0 - chr1 435840 435916 U2 0 - chr1 435842 435918 U4 0 - chr1 435842 435918 U4 0 - chr1 435844 435920 U2 0 - chr1 435844 435920 U2 0 - chr1 437189 437265 U2 0 + Could someone explain how U0, U1, U2 are assigned and what they are? Thanks,
U (unique) means the uniquely mapped reads. Maybe the script regards all the reads as unique reads.

And the integer means the number of mismatches.
__________________
Xi Wang

 11-23-2010, 09:16 PM #127 wdt Member   Location: Southwest Join Date: Oct 2009 Posts: 19 I have RNA-seq data analyzed using tophat that generated bam files for each sample. Each group (cases/controls) has 5 samples each. Would the following be correct way to use DEGseq 1. Convert BAMs to SAM to BED using samtools + sam2bed.pl 2. Use DEGseq samWrapper to test 5 samples in one group with 5 samples in the other to identify diff expressed genes? Thanks a lot! Last edited by wdt; 11-23-2010 at 09:20 PM.
11-23-2010, 09:21 PM   #128
Xi Wang
Senior Member

Location: MDC, Berlin, Germany

Join Date: Oct 2009
Posts: 317

Quote:
 Originally Posted by wdt I have RNA-seq data analyzed using tophat that generated bam files for each sample. Each group (cases/controls) has 5 samples each. Would the following be correct way to use DEGseq 1. Convert BAMs to BED using sam2bed.pl 2. Use DEGseq samWrapper to test 5 samples in one group with 5 samples in the other to identify diff expressed genes? Thanks a lot!
Agreed. But please note that you need first convert BAM to SAM using samtools.
__________________
Xi Wang

 11-29-2010, 11:19 AM #129 wdt Member   Location: Southwest Join Date: Oct 2009 Posts: 19 Many thanks for your quick replies about the DEGseq. Once BED files are provided, does DEGseq internally compute "raw counts" that are used for differential exp analysis? Is there a way to output those raw counts (or equivalent numbers) per sample? Thanks a lot!
11-29-2010, 05:38 PM   #130
Xi Wang
Senior Member

Location: MDC, Berlin, Germany

Join Date: Oct 2009
Posts: 317

Quote:
 Originally Posted by wdt Many thanks for your quick replies about the DEGseq. Once BED files are provided, does DEGseq internally compute "raw counts" that are used for differential exp analysis? Is there a way to output those raw counts (or equivalent numbers) per sample? Thanks a lot!
you can use the script below.

Code:
```refFlat <- "refFlat.txt"
mapResultBatch = c("sample1","sample2","sample3","...") # replace the file names accordingly
geneExpr <- "geneExpr.txt"   # you may specify the file name to save the gene expresion values
getGeneExp(mapResultBatch, refFlat=refFlat, output=geneExpr)```
__________________
Xi Wang

 12-06-2010, 08:03 AM #131 newbietonextgen Member   Location: USA Join Date: Nov 2010 Posts: 56 help With DEGseq Hello all, I have a 1.0 GB data file and was wondering how long it would take for the program to load this data? All i get after showing the path to sample A, is a spinning ball (mac) that keeps going on for half hour. I dont get the R prompt again and I just kill the process thinking some thing is wrong. Do i have to be patient ? The computer has 8 gb ram if that help. So please let me know. Sample bed format file using the samtobed script chr1 15562 15637 ILLUMINA-927B2F_0001:1:110:7901:1208#0/1 10 + chr1 15564 15636 ILLUMINA-927B2F_0001:1:92:5422:11873#0/1 10 + chr1 15564 15636 ILLUMINA-927B2F_0001:1:117:10103:16792#0/1 10 + chr1 16084 16159 ILLUMINA-927B2F_0001:1:3:3987:6468#0/1 10 - So please let me know if its the format or i just need the patience. Last edited by newbietonextgen; 12-06-2010 at 08:07 AM.
12-06-2010, 08:12 AM   #132
Xi Wang
Senior Member

Location: MDC, Berlin, Germany

Join Date: Oct 2009
Posts: 317

Quote:
 Originally Posted by newbietonextgen Hello all, I have a 1.0 GB data file and was wondering how long it would take for the program to load this data? All i get after showing the path to sample A, is a spinning ball (mac) that keeps going on for half hour. I just kill the process thinking some thing is wrong. Do i have to be patient ? The computer has 8 gb ram if that help. So please let me know. Thanks
What kind of data file you fed to DEGseq, BED, BAM? Usually, it couldn't need to take so much time to load 1GB data.
__________________
Xi Wang

 12-06-2010, 08:14 AM #133 newbietonextgen Member   Location: USA Join Date: Nov 2010 Posts: 56 Thanks Xi for the quick reply. It was a BED format file. I converted using the samTobed tools.
12-06-2010, 08:25 AM   #134
Xi Wang
Senior Member

Location: MDC, Berlin, Germany

Join Date: Oct 2009
Posts: 317

Quote:
 Originally Posted by newbietonextgen Thanks Xi for the quick reply. It was a BED format file. I converted using the samTobed tools.
I just saw you updated the message.
Were there any screen display?
__________________
Xi Wang

 12-06-2010, 08:31 AM #135 newbietonextgen Member   Location: USA Join Date: Nov 2010 Posts: 56 No. I have tried both formats: giving the path to the file and then setting up the working dir and then naming the file. I am using a 64 bit R and i am nots sure if it a problem with it. This is how the console looks: >library(DEGseq) Loading required package: qvalue Loading Tcl/Tk interface > sample A <- "path to the file (bed.txt)" | So there was no screen message after i hit return...
12-06-2010, 08:38 AM   #136
Xi Wang
Senior Member

Location: MDC, Berlin, Germany

Join Date: Oct 2009
Posts: 317

Quote:
 Originally Posted by newbietonextgen No. I have tried both formats: giving the path to the file and then setting up the working dir and then naming the file. I am using a 64 bit R and i am nots sure if it a problem with it. This is how the console looks: >library(DEGseq) Loading required package: qvalue Loading Tcl/Tk interface > sample A <- "path to the file (bed.txt)" | So there was no screen message after i hit return...
I found that you didn't use the most updated version of DEGseq.
http://bioconductor.org/packages/rel...ml/DEGseq.html

And second, in R, variables can't have space in them; And you should tell it where is your file, but not the sentence.
E.g.,
Code:
`sample_A <- "/home/username/data.bed"`
__________________
Xi Wang

Last edited by Xi Wang; 12-06-2010 at 08:43 AM.

12-07-2010, 07:10 AM   #138
Xi Wang
Senior Member

Location: MDC, Berlin, Germany

Join Date: Oct 2009
Posts: 317

Quote:
Hi,

Please show me your R script to run DEGseq. You can email me: wang-xi05@mails.tsinghua.edu.cn , if you don't want to put the details here.

Thanks.
__________________
Xi Wang

Last edited by Xi Wang; 12-11-2010 at 10:12 AM.

 07-21-2011, 04:36 AM #139 mgolo Member   Location: Denmark Join Date: Apr 2011 Posts: 10 DEGseq and expression of novel small RNAs Hi all! I´m new to the NGS business, and right now i have a lot of doubts about DE analysis. I have RNA-sequenced a bacterial transcriptome in 2 growth conditions, and I have 3 biological replicates for each condition: Condition A : Replicate 1A, Replicate 2A, Replicate 3A Condition B : Replicate 1B, Replicate 2B, Replicate 3B I have the bam an pileup files for each replicate. Now, my aim is compare the expression of non-annotated non-coding RNAs in my conditions A and B (so i will use a custom annotation file). I have read about DEGseq and i would like to use it for my DE analysis. But i have a number of questions about it: 1. What method would suit my analysis best? I have thought of using MARS... 2. How do I normalize my replicates? Should i use loess or median? What´s the difference between them? 3. What is better: to pool the 3 replicates of each condition or to analyze DE without pooling them? 4. Since my transcripts are not annotated i will have to use expression values based on raw read counts, right? Can i use the rawCount argument with the DEGseq function or is it only valid with the DEGexp function? If i use the MARS method is it automatically set to analyze raw counts? Thanks in advance for your help! Maria
07-21-2011, 06:50 PM   #140
Xi Wang
Senior Member

Location: MDC, Berlin, Germany

Join Date: Oct 2009
Posts: 317

Quote:
 Originally Posted by mgolo Hi all! I´m new to the NGS business, and right now i have a lot of doubts about DE analysis. I have RNA-sequenced a bacterial transcriptome in 2 growth conditions, and I have 3 biological replicates for each condition: Condition A : Replicate 1A, Replicate 2A, Replicate 3A Condition B : Replicate 1B, Replicate 2B, Replicate 3B I have the bam an pileup files for each replicate. Now, my aim is compare the expression of non-annotated non-coding RNAs in my conditions A and B (so i will use a custom annotation file). I have read about DEGseq and i would like to use it for my DE analysis. But i have a number of questions about it: 1. What method would suit my analysis best? I have thought of using MARS... 2. How do I normalize my replicates? Should i use loess or median? What´s the difference between them? 3. What is better: to pool the 3 replicates of each condition or to analyze DE without pooling them? 4. Since my transcripts are not annotated i will have to use expression values based on raw read counts, right? Can i use the rawCount argument with the DEGseq function or is it only valid with the DEGexp function? If i use the MARS method is it automatically set to analyze raw counts? Thanks in advance for your help! Maria
Hi Maria

1&2. The methods for DEG detection and the normalization beforehand should depend on how your data distributed. You may try all of them and choose the best one.

3. For biological replicates, it's better not to pool them together.

4. Raw read counts have nothing to do with gene annotation. In our documents, the opposite of 'raw read counts' is RPKM vaules. For the unannotated non-RNAs, you'd better analyze the gene structure first and then the DEGs.

Btw, we are working a new version of DEGseq, which will be more suitable for biological replicates.
__________________
Xi Wang

 Tags degseq, rna-seq