October 2012 - galaxy-user - lists.galaxyproject.org

cuffdiff variance calculations
by i b 17 Oct '12

17 Oct '12

Hello forum, I was reading some of the theory behind cuffdiff and how it calculates for DE genes. My question: if I do not have replicates, how does cuffdiff do the test statistic which is dependent on the variance? The manual says "Note that in order to calculate the test statistic T, we need to know the variance of the expression level in each condition" Doesn't this mean that I am supposed to have replicates?Otherwise how can we have variance if we only have ONE SAMPLE? I would appreciate if anyone could explainthis to me. Ciffdiff caclulates DE even when I give him one sample agaist another one...how reliable is the reult? Thanks... ib

1 0

encode rna-seq data
by laura lafontaine 17 Oct '12

17 Oct '12

Hi everyone, I'm new in Galaxy and I want to know if is possible download whole rna-seq data from ENCODE project to my local galaxy instance (I have a disk of 5 To) . If yes, would you explain me how I can do this? Thank you in advance for your answer, L

2 1

SNP Calling
by Francesco Vitiello 17 Oct '12

17 Oct '12

Does anyone knows a step-by-step pipeline to SNP calling on illumina dataset? >From the alignment to the end.... thanks Francesco --

1 0

Galaxy problem
by Francesco Vitiello 17 Oct '12

17 Oct '12

Dear all, I' really new to the use of Galaxy, NGS analysis and SNP calling. I have a problem. After that I generate pileup and filter some reads with the SAMtools filter pileup The resulting data set contain about 2200 SNPs At least I compare my result data set whit dbSNP132.txt and 1kg.lc.2010_7.CEU.liftedhg19.pgSnp. The resulting data set contain 700 SNPs. Now the question: The problem is that I must have 200 SNPs instead of 700.... I have check the resutls and I saw that some result are duplicated for all parameters but had a different rs ID. how can I remove the wrong one??? please help me. some one I heard about SOAPsnp but I don't know how to use. Could anyone please help me. Thanks Francesco --

1 0

cuffdiff fpkm=0 "not significant"
by i b 17 Oct '12

17 Oct '12

Dear forum, whoever knows this please tell me!!!!! Given two fpkm . e.g. fpkm 1=2922828 and fpkm 2=0, why cuffdiff does not calculate differential expression. those genes are listed in cuffdiff as not significant and the log2 fold change is infinite , either negative or positve...how do i consider them when i want to pull out the genes that are differentially expressed between treated and untreated samples? how do i interpret these data? thanks, ib

2 1

cuffdiff results interpretation: high fold change but not significant
by Qian Dong 16 Oct '12

16 Oct '12

Dear team, I've been having some problem with cuffdiff results interpretation. I understand that on for a certain sample, it has to have a log2(fold change) that is larger than the cutoff and also have a q-value that smaller than a cutoff. I have a lot of genes that have a really high fold change and also high q-values so they are 'not significant'. My questions will be: 1. Since they are 'not signaficant', should I not take those genes into consideration even if they have high fold change? 2. Some of them have really low value_1 or value_2 (less than 1). Is this why they have a high q-value? 3. How can I get a high q-value if both value_1 and value_2 are quite large and the fold change is also large? The reason I ask those questions is in my cuffdiff results, I found some really interesting genes of high fold change but they are 'not significant' . I am really not sure whether I have to give they up because of the high q-value. Thank you very much, Qian -- Qian Dong Bauer Lab, MCBD Simon Hall: 313-317 212 S. Hawthorne Dr. Bloomington, IN 47405 Email:dong3@indiana.edu Lab Phone:812-855-8443

1 0

"intersect" operation
by FNU Ujani Chakraborty 16 Oct '12

16 Oct '12

Hi! I was trying to intersect two data sets yesterday and today....and it seems like in the dialog box that asks me to enter the names of the 2 datasets that I wish to intersect, the order in which I list the two datasets makes a difference and gives me different number of overlaps. So if I have two datasets A and B if I enter A first and B second in the dialog box, the result is different from if I enter B first and A second in the dialog box. I dont understand how this is possible, since the overlap should not depend on the order of the entries. Am I doing or understanding something wrong? Any help will be appreciated! Ujani

2 1

Upload data to local Galaxy by copying them to the file folder
by Jerzy Dyczkowski 13 Oct '12

13 Oct '12

Hello, How to put data to the local Galaxy server via ftp, ssh or copying directly to some subfolder in galaxy-dist? We installed a local Galaxy instance and it is running. Manual uploading files from user PCs is slow due to less than perfect connections in the local network. We would prefer uploading large files by ssh from the central cluster to the server on which Galaxy is running. I can conect to the Galaxy server and Galaxy folder structure, however, I cannot find where to put files. Are they all renamed and coded as something like: galaxy-dist/database/files/000/dataset_38.dat ? The option of uploading via ftp on main.g2.bx.psu.edu is absent in local Galaxy distribution. best regards, Jerzy

2 1

job information on galaxy
by i b 12 Oct '12

12 Oct '12

Hi, I am trying to open the "i" icon on galaxy to look at the information of each job, but this does not opne for old jobs, only the last recent one ... Any suggestion? Thanks, ib

3 2

Error message on local install
by Chris Merrikh 12 Oct '12

12 Oct '12

Hi, I'm trying to solve an issue I'm having with my local installation of Galaxy (installed on my own computer, rather than on a server). I'm using data in the form of fastq files from an Illumina Hi-seq and I want Galaxy to parse the bar coded sequences out into individual files for me. I've been using the public server in the past, and I'm able to use the Groomer, Joiner, Reverse-Complement, Trimmer, and Barcode Splitter tools just fine. Now I'm trying to do the same thing locally on the same files. Both the large file (38 GB) and a smaller file consisting of the first 10k lines will upload just fine. However, I can only get the Groomer to work on the small file. When I use it on the large file I get an error: "Error executing tool: maximum recursion depth exceeded while calling a Python object." Any help on this would be greatly appreciated! - Chris M.

3 2