Problems getting started with RSEM

[j.mi...@ec.gc.ca]

Hello, I am trying to get started with RSEM (1.2.4) and also I am fairly new to using LINUX so please keep that in mind, thanks.

I am trying to load and run RSEM and I believe that I have loaded the program properly but when I try to do rsem-prepare-reference I get:

bash: rsem-prepare-reference: command not found

I wonder if I did the install correctly, because one thing that strikes me is that was no "complete" or "DONE" type of message when I ran "make" on rsem and there were a few warning, but no errors. I noticed on other programs that I compiled from source code that the last output line gave some sort of verification that the install was successful but I did not see that on rsem-1.2.4 (or on rsem-1.2.3 which I tried to install a few weeks ago before getting side-tracked onto other things).

Here is the final few lines of the output from running make (full output is appended to this meesage)

g++ -O3 -Wall samValidator.cpp sam/libbam.a -lz -o rsem-sam-validator

utils.h:26: warning: ‘verbose’ defined but not used

g++ -O3 -Wall scanForPairedEndReads.cpp sam/libbam.a -lz -o rsem-scan-for-paired-end-reads

utils.h:26: warning: ‘verbose’ defined but not used

[adminrig@JR09110715 rsem-1.2.4]$

My OS is Linux CentOS 5.4, x86_64 arcitecture and I did put both RSEM and bowtie in the environments PATH variable using this command:

PATH=$PATH:"$home/adminrig/RSEM/rsem-1.2.4

PATH=$PATH:"$home/adminrig/Bowtie/bowtie-1.0.0

And confirmed this using the printenv PATH command

Running 'make' in EBSeq seemed to work fine based on the output (I have not actually used the program yet since I need to use RSEM first). Here are the last few lines. Again I had some "defined but not used" warnings. Full output is also in the appended file.

** building package indices

** installing vignettes

** testing if installed package can be loaded

 

* DONE (EBSeq)

g++ -O3 -Wall calcClusteringInfo.cpp -o rsem-for-ebseq-calculate-clustering-info

[adminrig@JR09110715 EBSeq]$

I also installed or updated the following programs / utilities prior to building rsem:

gcc-c++  4.1.2-54.el5

perl 5.8.8-38.el5

R 2.15.2-1.el5

ncurses-devel 5.5-24.20060715

Being a rookie in this area I assume I am missing something basic but I have tried everything I could think of trying before asking here. Thank-you for you time and I look forward to any help you might be able to provide.

Jason Miller

Aquatic Contaminants Research Division

National Water Research Institute
Environment Canada - Canada Centre for Inland Waters
867 Lakeshore Rd.
Burlington, Ontario, Canada

[Colin Dewey]

Hi Jason,

That error suggest that either your PATH variable is not set up correctly, or that the rsem-prepare-reference script has somehow gone missing in your install, or has lost its executable permissions.  To check your PATH, perhaps try running the command "rsem-preref" (which I can see was built in your make log) and see if it gives you a command not found.  If so, then check the construction of your PATH variable.  To check the rsem-prepare-reference script, perhaps try listing it with 

ls -l $home/adminrig/RSEM/rsem-1.2.4/rsem-prepare-reference

and check that it both exists and has its executable permissions set appropriately for the user you are trying to run as.

Hope that helps,

Colin

--
RSEM website: http://deweylab.biostat.wisc.edu/rsem/
---
You received this message because you are subscribed to the Google Groups "RSEM Users" group.
To unsubscribe from this group and stop receiving emails from it, send an email to rsem-users+...@googlegroups.com.
To post to this group, send email to rsem-...@googlegroups.com.
Visit this group at http://groups.google.com/group/rsem-users?hl=en-US.
 
 
<rsem 1 2 4 make results.pdf>

[j.mi...@ec.gc.ca]

Thank-you Colin for the reply. It turned out to be a problem with the PATH variable as you suggested (I ended up rebuilding the rsem and bowtie files in /usr/local/bin and then placed them into the PATH variable and that did the trick).

Fingers crossed that things go smoothly from here. Thanks again.

Jason

[j.mi...@ec.gc.ca]

Hello again, sorry to bother you all with more newbie questions but I am having a problem know with rsem-calculate-expression. I am getting an error 

[samopen] no @SQ lines in the header

[sam_read] missing header? Abort!

I tried to find some answers to this problem already but the problem is I don't understand what files should contain the @SQ lines.

Here's what I did so far

ran rsem-prepare-reference using a mouse cDNA fasta file downloaded from ensembl as the 'reference genome':

rsem-prepare-reference Mus-musculus.GRCm38.71.cdna.all.fa May1_Trial1ref

where Mus-musculus.GRCm38.71.cdna.all.fa is the cDNA FASTA file obtained from ensembl

and May1_Trial1ref is the reference name

that appeared to run just fine and created the expected files, so then I tried rsem-calculate-expression

(for the record I am trying to use the Table S1_1 simulated data just to get a feel for the program)

My command was:

rsem-calculate-expression --no-qualities /simulation_data_ensembl/table_S1/table_S1_1.fa May1_Trial1ref Trial1sample

I have rsem and bowtie in the PATH variable

any ideas?

Thanks again

[Colin Dewey]

Hi Jason,

First, if you can provide the complete terminal output after you run the rsem-calculate-expression command, that would help in diagnosing what is wrong here.  rsem-calculate-expression internally runs samtools to convert a SAM output from Bowtie into a BAM file; the error messages you report are likely from this process.  My guess is that bowtie is not running correctly.  Are you able to run the "bowtie" command manually?

Colin

[j.mi...@ec.gc.ca]

Thanks again for the swift reply. I decided I would re-run rsem-prepare-reference and rsem-calculate-expression before replying (to copy all the outputs and also show which files are created at each step). I first created a new file first and copied the two FASTA files (Mus-musculus.GRCm38.71.cdna.all.fa and table_S1_1.fa = "reference" and "sample"). I did this simply for clarity. rsem-prepare-reference proceeded seemingly identical to last time, but this time rsem-calculate-expression also worked! For the record I did make more than a few attempts to get rsem-calculate-expression to run before posting here, I guess there was some issue with the files or maybe I had errors in the file names?

Thanks

Jason

[b...@cs.wisc.edu]

Hi Jason,

I'm not sure what problem it might be. Your original error message suggest
that samtools cannot read the header information from the BAM file. Since
it works this time, maybe you can continue with downstream analysis.

Best,
Bo

> Thanks again for the swift reply. I decided I would
> re-run rsem-prepare-reference and rsem-calculate-expression before
> replying (to copy all the outputs and also show which files are created at
> each step). I first created a new file first and copied the two FASTA
> files
> (Mus-musculus.GRCm38.71.cdna.all.fa and table_S1_1.fa = "reference" and
> "sample"). I did this simply for clarity. rsem-prepare-reference proceeded
> seemingly identical to last time, but this time rsem-calculate-expression
> also worked! For the record I did make more than a few attempts to
> get rsem-calculate-expression to run before posting here, I guess there
> was
> some issue with the files or maybe I had errors in the file names?
>
> Thanks
>
> Jason
>
>
> On Thursday, 2 May 2013 12:36:29 UTC-4, Colin Dewey wrote:
>>
>> Hi Jason,
>>
>> First, if you can provide the complete terminal output after you run
>> the rsem-calculate-expression command, that would help in diagnosing
>> what
>> is wrong here. rsem-calculate-expression internally runs samtools to
>> convert a SAM output from Bowtie into a BAM file; the error messages you
>> report are likely from this process. My guess is that bowtie is not
>> running correctly. Are you able to run the "bowtie" command manually?
>>
>> Colin
>>

>>>> *Jason Miller*

>>>>
>>>> Aquatic Contaminants Research Division
>>>> National Water Research Institute
>>>> Environment Canada - Canada Centre for Inland Waters
>>>> 867 Lakeshore Rd.
>>>> Burlington, Ontario, Canada
>>>>
>>>>
>>>>
>>>> --
>>>> RSEM website: http://deweylab.biostat.wisc.edu/rsem/
>>>> ---
>>>> You received this message because you are subscribed to the Google
>>>> Groups "RSEM Users" group.
>>>> To unsubscribe from this group and stop receiving emails from it, send
>>>> an email to rsem-users+...@googlegroups.com.
>>>> To post to this group, send email to rsem-...@googlegroups.com.
>>>> Visit this group at
>>>> http://groups.google.com/group/rsem-users?hl=en-US.
>>>>
>>>>
>>>> <rsem 1 2 4 make results.pdf>
>>>>
>>>>
>>>>
>> --
>> RSEM website: http://deweylab.biostat.wisc.edu/rsem/
>> ---
>> You received this message because you are subscribed to the Google
>> Groups
>> "RSEM Users" group.
>> To unsubscribe from this group and stop receiving emails from it, send
>> an

>> email to rsem-users+...@googlegroups.com <javascript:>.

>> To post to this group, send email to

>> rsem-...@googlegroups.com<javascript:>
>> .