Question: CuffDiff error, "Uncaught exception in exposed API method," job runs, but no results
0
gravatar for igboin.2
3.8 years ago by
igboin.20
United States
igboin.20 wrote:
Hello,
I’ve run into a problem while analyzing some RNAseq data on usegalaxy.com. I am looking at bacterial and host gene expression during infection. The host is Drosophila and the bacterium is F. nucleatum ATCC 10953 (Fn). I made and sequenced 3 RNA libraries: Fn alone (RNA from in vitro grown bacteria), Drosophila alone (RNA from uninfected flies), Fn/Drosophila mix (RNA from Fn infected flies). 
I aligned the Drosophila reads with Tophat and performed differential analysis (Drosophila alone vs. Drosophila/Fn mix) using Cuffdiff, with no problem. 
Problem: I aligned the Fn reads with Bowtie, but when I try to perform differential analysis using Cuffdiff (Fn alone vs. Drosophila/Fn mix) I keep getting errors as follows:  
 
1st time: I ran CuffDiff on BAM files generated by Bowtie (I didn't filter out unmapped reads). 
I got the following error in red in the History: “An error occurred while getting updates from the server. Please contact Galaxy administrator if the problem persists.” 

I clicked on “details,” and saw the following (this is not the original, just the most recent one, dated Sept 3rd. I’ve run the analysis multiple times since August when it first happened and gotten the same error): 
 
user  username    christinaigboin

quota_percent    55

total_disk_usage    147709821229

nice_total_disk_usage    137.6 GB

email    igboin.22osu.edu
is_admin    false

tags_used    

model_class    User

id    a175da8e4b898694

source    HistoryCOntents(7f397613db43f024,270)

xhr   readyState    4

responseText    {"err_msg": "Uncaught exception in exposed API method:", "err_code": 0}

responseJSON  err_msg    Uncaught exception in exposed API method:

                           err_code    0

status     500

statusText    Internal Server Error

responseHeaders    Server    nginx/1.4.7

Date    Tue, 03 Sep 2014 00:08:34 GMT

Content-Type    application/json

Transfer-Encoding    chunked

Connection    keep-alive

Cache-Control    max-age=0,no-cache,no-store

options  data    

parse    true


The job still ran but every gene got a “NO TEST” under status and “0” under value_1 and value_2 (FPKM for each of the two samples/conditions being compared). 
 
2nd time: Thinking it was a server issue, I repeated the exact same analysis. Again I got the error in red, the job ran, but every gene got a “NO TEST” under status and “0” under value_1 and value_2.
 
3rd time: I filtered out the unmapped reads from the Bowtie output, and repeated the analysis on BAM files containing only mapped reads. Again I got the error in red, the job ran, but every gene got a “NO TEST” under status and “0” under value_1 and value_2.
 
4th time: I changed the header of the fasta file, which was originally: “>gi|254302039|ref|NZ_CM000440.1| Fusobacterium nucleatum subsp. polymorphum ATCC 10953 chromosome, whole genome shotgun sequence”, to: “>NZ_CM000440.1”, in order to match what is written in the .gff file, under the Seqid column for each gene (“NZ_CM000440.1”). I performed the analysis using this updated fasta file and the BAM files containing only mapped reads. Again I got the error in red, the job ran, but every gene got a “NO TEST” under status and “0” under value_1 and value_2.
 
5th time (ran overnight Sept 2-3): I reran the analysis using the updated fasta file and SAM files containing only the mapped reads (did a BAM to SAM conversion). Again I got the error in red, the job ran, but every gene got a “NO TEST” under status and “0” under value_1 and value_2.
 
The Cufflinks page says that Cuffdiff can be used with RNA-seq data from bacteria (http://cufflinks.cbcb.umd.edu/faq.html#bact), so I am not sure what I am doing wrong.   
I appreciate any help you can give me.
Thanks,
Christina
rna-seq methods bacteria cuffdiff • 1.5k views
ADD COMMENTlink modified 3.8 years ago by Jennifer Hillman Jackson25k • written 3.8 years ago by igboin.20
0
gravatar for Jennifer Hillman Jackson
3.8 years ago by
United States
Jennifer Hillman Jackson25k wrote:

Hello,

This result means (from the manual http://cufflinks.cbcb.umd.edu/manual.html):

 NOTEST (not enough alignments for testing)

I would go back and review the input sequences for quality, the overall mapping rate from Bowtie, then once all that checks out and the rates are still low or your data is just simply sparse intrinsically - see the google group tophat.cufflinks@gmail.com and perhaps seqanswers.com and biostars.org for similar questions about the best parameters when exploring bacterial genomes. And while this pipeline is used, it is not the favorite one. You'll see the other recommendations in these posts. I believe that most of the tools are in the Tool Shed, which are appropriate for a local or cloud Galaxy.

http://wiki.galaxyproject.org/BigPicture/Choices
http://usegalaxy.org/toolshed

Best,

Jen
Galaxy team

ADD COMMENTlink written 3.8 years ago by Jennifer Hillman Jackson25k

Thanks, Jen, I'll do what you recommend.

ADD REPLYlink written 3.8 years ago by igboin.20
Please log in to add an answer.

Help
Access

Use of this site constitutes acceptance of our User Agreement and Privacy Policy.
Powered by Biostar version 16.09
Traffic: 99 users visited in the last hour