Segmentation fault (core dumped) in genomegenerate step

[JAYRJPT]

I am using a customized human reference genome containing only specific portion of chromosome 22 and 9. Total size of the reference genome is ~4kb. while generating the index I am getting the following error-

Feb 03 20:03:01 ..... started STAR run
Feb 03 20:03:01 ... starting to generate Genome files
Feb 03 20:03:01 ... starting to sort Suffix Array. This may take a long time...
Feb 03 20:03:01 ... sorting Suffix Array chunks and saving them to disk...
Feb 03 20:03:01 ... loading chunks from disk, packing SA...
Feb 03 20:03:01 ... finished generating suffix array
Feb 03 20:03:01 ... generating Suffix Array index
Feb 03 20:03:01 ... completed Suffix Array index
Feb 03 20:03:01 ..... processing annotations GTF
**Segmentation fault (core dumped)**

Here is my command- STAR --runMode genomeGenerate --genomeDir //STR_INDEX --genomeFastaFiles 22_9.fasta --genomeChrBinNbits 8 --genomeSAindexNbases 5 --sjdbGTFfile gencode.v36.annotation_edited.gtf

I have read the earlier post and tried the suggestion but could not able to troubleshoot the issue.

Kindly help to troubleshoot the issue.

Thanks and Regards,

JAY

[alexdobin]

Hi Jay,

Does it work when you generate index with the standard FASTA and GTF? If so, the first thing to check is whether the formatting of GTF file is correct. Did you change to coordinate of exons in the GTF accordingly when you selected a small portion of the chromosome?

Cheers Alex

[JAYRJPT]

Hi Alex, Yes, it is working with standard FASTA and GTF. No I didn't change the coordinate of the exons in the GTF file rather I have used the standard GTF file for generating index for the edited reference FASTA.

[alexdobin]

Hi Jay,

you would have to change the exon coordinates so that agree with the new genome sequence. For instance, if you removed 10,000 bases from the beginning of chr1, all chr1 exons in the GTF should have 10,000 subtracted from their start and end positions.

Cheers Alex