Segmentation fault on first-pass mapping from BAM files

[Leonor Palmeira]

Dear Stars,

I use STAR with fastq.gz files without any problem, but for a new experiment, I only have BAM files and no fastqs.
I tried using STAR with appropriate changes but get a "Segmentation fault", so I guess I am doing something wrong.

This is the output I get:

/var/spool/slurm/job514924/slurm_script: line 99:  4975 Segmentation fault      $STAR --runThreadN $ncores --outSAMattributes All --genomeLoad NoSharedMemory --outSAMstrandField intronMotif --genomeDir $GLOBALSCRATCH/genomes/homo_sapiens/hg19/star_hash --runMode inputAlignmentsFromBAM --bamRemoveDuplicatesType UniqueIdentical --inputBAMfile $datadir"/"$inputbam

I can send you the Log.out which ends with:
Apr 23 15:16:38 ..... Reading from BAM, remove duplicates, output BAM

Could you please help me figure out the issue?
Thanks a lot for your help
Leonor

[Alexander Dobin]

Hi Leonor,

you are trying to map the reads recorded in a BAM file - this feature is not implemented yet. --runMode inputAlignmentsFromBAM only works for post-map processing of the BAM files, such as wiggle generation or duplicate removal. You need to extract fastqs from the BAM files using, for instance, bedtools (http://bedtools.readthedocs.org/en/latest/content/tools/bamtofastq.html) or picard (http://broadinstitute.github.io/picard/command-line-overview.html#SamToFastq)

Cheers

Alex