STAR aligner Index build
Hi, I am trying to build STAR index with the following commands, both the Gtf and fasta files were downloaded from Gencode. Following is the commands been used.
STAR --runMode genomeGenerate --genomeDir /data/WHRI-GenomeCentre/Genome/Human/hg19/STAR_index/Genome_14.2.18 --genomeFastaFiles GRCh38.p10.genome.fa --sjdbGTFfile gencode.v27.annotation.gtf --sjdbOverhang 74 --runThreadN 8
The command is working fine and creating .txt files and later aborting. I cannot figure out what the reason is. Below is the error message. error : job_scripts/219671: line 26: 28981 Killed
I also checked the log.out file and the program is terminated after the below warning:
WARNING: long repeat for junction # 255126 : chr15 88855522 88855578; left shift = 14; right shift = 255
Feb 14 12:48:27 Finished preparing junctions
Feb 14 12:48:27 ..... inserting junctions into the genome indices
Can someone help me with what is the reason for the error. Thanks in advance Hothri
• 122 views
•
link
updated
by
GenoMax
•
written
by
hothriananya
•
written
by
hothriananya
0 answers
No answers yet.
Log in to answer this question.
More posts like this
-
STAR Genome index Error
written by Prasanna •
-
Help with for loop for sequence alignment using STAR
written by Amaranta Remedios •
-
STAR Generating Genome Indices error
written by Balatheskulter •
-
RSEM: transcripts.fa.grp not found (should it have been generated?)
written by jhanks1981 •
-
no input BAM records found in ./tophat_out_SRR5481725/tmp/right_kept_reads.m2g_um.candidates15.bam …
written by ALI Hkz •
-
STAR index doesn't finish
written by alvarocentron91 •
-
STAR index fail with gtf, works with gff3
written by Bastien Hervé
-
stringtie cannot recognize Gencode GTF
written by dec986
-
TopHat Failed In 'Reporting output tracks' Stage
written by ivan.molodtsov •
-
Tophat failed at 'Mapping left_kept_reads to transcriptome'
written by biocyberman
That is only a warning. It may not be related to the program getting killed.
Are you assigning enough RAM/wall clock time for this job (if you are using a scheduler). How much RAM do you have available? STAR requires 30+G of free RAM for human genome.
yes, I have more than 50+ GB RAM available and also clock tie for the job is 24hrs
Ok. Enough space available in /tmp and the location where you are writing the indexes out to? Did you build STAR yourself or are you using supplied binaries ( https://groups.google.com/forum/#!topic/rna-star/c4ujBEDOwGo )?
Yes I have got 1TB space. No, I havent build the STAR. I will check with the IT to see what binaries have been used to build STAR. Thank you.
Alex Dobin has pre-made STAR indexes available here. In case you are able to use them.
I can give a try, but I would prefer to build indexes. Thanks for the suggestions.
are the chromosome names the same between the annotation file and the genome one? alternatively it could be a memory problem, how much RAM do you have?
Both the annotation and genome files were downloaded from Gencode current release.