BRAKER for large genome?

Question

[Dave]

I have a genome assembly of 2.1 GB, and I'm working on annotation with BRAKER2. I have lots of RNAseq data, and I was hoping to use that, as well as the OrthoDB arthropod protein set for gene prediction in --etpmode mode. However, samtools threw the error regarding the size limit for BAI indexes.

[E::sam_index] Read 'HWI-D00381:362:C7HLTANXX:4:1310:18191:84687' with ref_name='Scaffold_49954', ref_length=559598383, flags=417, pos=536862089 cannot be indexed
samtools index: failed to create index for "/home/FCAM/dangelini/data/short.reads/PK_SRvsChr.bam": Numerical result out of range

I am currently running samtools index -c to create a CSI index. However, I'm concerned about the needs of BRAKER (running v2.1.6) in terms of the BAM index. -- Does BRAKER even need an indexed BAM file? I cannot find clear docs on this issue.

Thanks!

Answer 1

[Buffo]

It seems like a mapping error, the same issue was reported years ago here.