[E::bwa_idx_load_from_disk] fail to locate the index files

[V-Catherine]

Hello Jeremiah, I have an error while trying to run SvABA, similar to the one Tiffany had but that does not seem to be linked. I have run the following command : svaba run -t sid_t_genome_hg38.bam -n sid_n_genome_hg38_20190127.bam -p 1 -a somatic_run -G Homo_sapiens_assembly38.fasta And I get the following error message: [E::bwa_idx_load_from_disk] fail to locate the index files ERROR: Unable to open index file: Homo_sapiens_assembly38.fasta I have checked the bams: they were aligned with the same genome reference than the one I use to run SvABA. I have also tried to re-indexed the reference genome first with bwa and then with samtools and none of the reindexing solved the error. Do you have any idea what the problem may be ? Thanks in advance for you help Catherine

[walaj]

Hi Catherine, Hmm, this is odd. Would you be able to email me ([email protected]) a Dropbox link with the reference genomes and indicies? I want to try and recreate so I can see what is going on. Jeremiah

[V-Catherine]

Hi, Thanks for your answer. The reference genome I use is GATK's hg38. As for the genomes I'm working on I need to check if I'm allowed to share them and how. I'll come back to you next week. Thanks again Catherine

[walaj]

Hi Catherine, I'm still not able to recreate this error. I indexed hg38 and it loads well. Are you also indexing with "samtools faidx"? Best, Jeremiah

[V-Catherine]

Hi Jeremiah, Thanks for your message. I re-indexed first with bwa (bwa index Homo_sapiens_assembly38.fasta) but I got the same error. This is why I then also indexed with samtools (samtools faidx Homo_sapiens_assembly38.fasta) and I still got the same error. I did not want to bother you with this issue until I have tried a few other things with another installation of svaba and another batch of samples but I won't be able to do this before one or two more weeks. That's why I did not come back to you but I will do so as soon as I have some more information to share on this issue :-)

[osowiecki]

Same issue here.

---

--- Running svaba SV and indel detection on 20 threads --- --- (inspect *.log for real-time progress updates) ---

[E::bwa_idx_load_from_disk] fail to locate the index files ERROR: Unable to open index file: ./genome.fa

EDIT : creating another bwa index fixed the issue

[nandobonf]

Same problem here. I am using fasta and index file from GATK bundle.

[phansol]

I had the same error. I guess it requires not only .fai index but other indices. In my case, this error was solved by creating bwa index (.amb, ann, bwt, etc.) using the following command. bwa index -a bwtsw [reference.fa]

[nukaemon]

I think SvABA expects the bwa index files to be named as suffixed in a strict way, that is, genome.fa.amb, genome.fa.ann, genome.fa.bwt, genome.fa.pac, and genome.fa.sa if the file given to -G option is genome.fa. I usually use -p option of bwa-index so that a series of the index files are named in genome.amb and so on, but this was the cause of the error in my case. So, regenerating bwa index files is partially correct, but do not use -p option.

[walaj]

Ah-ha! Thank you for solving this issue @Hansol-Park and @nukaemon. I'm leaving this open as a reminder to patch this so that either naming convention can work.

[bdabykov]

I solved indexing the fasta files to align. In my case indexing the contigs.fasta. like this: bwa index contigs.fasta and then running the align. bwa mem -P contigs.fasta SRR1.fastq SRR2.fastq> align_1sam