This is a test version of Biostars. For the public version, visit https://www.biostars.org.
STAR on Galaxy

Hello,

I have been having issues trying to implement an RNA-seq pipeline through Galaxy using an AWS cloud instance. My goal is to start from raw reads and get to gene counts. However, I have been having issues getting STAR to run. My current error is 

Fatal error: Matched on fatal error
EXITING: fatal error trying to allocate genome arrays, exception thrown: std::bad_alloc
Possible cause 1: not enough RAM. Check if you have enough RAM 32002674832 bytes
Possible cause 2: not enough virtual memory allowed with ulimit. SOLUTION: run ulimit -v 32002674832
Jun 03 21:16:00 ...... FATAL ERROR, exiting

I don't believe this is a memory issue as my current cloud instance is using a .xlarge image and the cloudman console says I have used 43G of my total 100G of data. My input datasets are a pair of paired end reads, loaded as fastq.gz. I have also loaded USCS GR38 gene annotation to be used and match the built in reference genome hg38 (which is from USCS I believe.

My STAR input is 

-paired-end as individual datasets.
-use built in index
    -use genome reference without builtin gene-model
    -Human (Homo sapiens) (b38): hg38
    -Gene model: UCSC_GR38_annotation.gtf
-length of genomic sequence around annotated junctions: 49

(read lengths were 50)

any help would be appreciated

rna-seq
• 3,074 views • link
written by jmgoldstein7 • jmgoldstein7

Memory is not storage. How much RAM do you have available in your instance? STAR is kinda known to need a lot of that.

• 0 views • link
written by cschu181 cschu181

hi jmgoldstein7,

I added (code) markup to your post for increased readability. You can do this by selecting the text and clicking the 101010 button. When you compose or edit a post that button is in your toolbar, see image below:

101010 Button

• 0 views • link
written by lieven.sterck lieven.sterck

If you are really sure you have the necessary amount of RAM available, did you try the proposed solution?

SOLUTION: run ulimit -v 32002674832

• 0 views • link
written by h.mon h.mon

1 answer

The m5.xlarge instance that I am using has 16 GB of memory. And from the line 

Check if you have enough RAM 32002674832 bytes

Which equates to 32 GB I think you are right and I was mixing up memory and storage like an amateur! Thank you, and for the tip about readability. For future reference, if I were to try the proposed solution, on Galaxy it is obviously a little different, is the section

Maximum available RAM (in bytes) for sorting
0
If 0, this will be set to the genome index size. This is typically only changed in cases where an error is produced. (--limitBAMsortRAM)

The right section to change to the memory limit? Thanks again

• 0 views • link
written by jmgoldstein7 • jmgoldstein7

Log in to answer this question.

More posts like this