Genome generate STAR

[sphil]

Hi
I wanted to generate my own genome using STAR. It started yesterday around 4 PM (CET) and hasn't finished yet. Any ideas how long it should take?

Command line was: ./STAR_2.3.0e/STAR --runMode genomeGenerate --genomeDir ./star_hg19_ref --genomeFastaFiles ./ --sjdbGTFfile ./gene_splits.gtf --runThreadN 6

std.out was:

Jul 10 15:55:16 ..... Started STAR run
Jul 10 15:55:16 ... Starting to generate Genome files

Maybe it broke?
Any suggestions?

Thanks,

Phil

[dpryan]

I don't recall how long my last run of that took, but it certainly didn't take overnight! You might check to see if it's actually using resources or if something else is already using all of the CPU.

[sphil]

nope nothing else is running. It's still running. It does not use 6 threads though but... i think i'll download the provided version now.... . The one I used was the normal UCSC verison. Dunno why it does not work...

[bruce01]

Have you made the genomeDir? Also maybe specify the fasta files. I remember I had a little difficulty until I spoon-fed it.

[alexdobin]

Hi Phil,

you need to list all the .fasta files in the in the command line, not just the directory name
--genomeFastaFiles g1.fa g2.fa ...
You can try to use wildcards:
--genomeFastaFiles ./*.fa

Also, for generating genome with annotations you need to specify
--sjdbOverhang <L>
where <L> is ideally read (mate) length -1 - you can also set it generically at 100.

The run should take a few hours. If fasta files were processed without a problem, within the first few minutes it it should get to the " ... starting to sort Suffix Array. This may take a long time ..." message .

Cheers
Alex

[sphil]

Thanks a lot, will give it a try!

[natallah]

Issues with STAR genomeGenerate on hg19

I am trying to generate genome files for hg19 (using hg19 annotations downloaded from UCSC). I have to use this genome build, however I am having trouble with the genomeGenerate command. I have used STAR in the past fine with the ENSEMBL genome and gtf file, so I am not sure why I am having problems now. It has been running for over 24 hours now and seems stalled. I cannot find pre-build genome files for hg19 that have annotations. The log file is attached.

The command I am using is:

--runMode genomeGenerate --genomeDir ./STARoutput_1 --genomeFastaFiles genome.fa --sjdbGTFfile genome_ref_notID.gtf --sjdbOverhang 99 --runThreadN 20