bbmap reference mapping error

Question

[drikaul]

Hi!

I'm trying to map my genomes to the C_glabrata reference genome and I'm having some issues with the backend java. I'm running bbmap.sh (Version 38.58) using the following command:

bbmap.sh ref=C_glabrata_CBS138_current_chromosomes.fasta in1=<R1.fastq> in2=<R2.fastq> outm=output_sam/<sampleName>_mapped.sam trd=t -Xmx1g

and get the following error:

Retaining first best site only for ambiguous mappings.
Writing reference.
Executing dna.FastaToChromArrays2 [C_glabrata_CBS138_current_chromosomes.fasta, 1, writeinthread=false, genscaffoldinfo=true, retain, waitforwriting=false, gz=true, maxlen=536670912, writechroms=true, minscaf=1, midpad=300, startpad=8000, stoppad=8000, nodisk=false]

Set genScaffoldInfo=true
Writing chunk 1
Exception in thread "main" java.lang.AssertionError: The index format has changed in this version of BBTools.  Please delete the /ref/ directory and re-index from the reference fasta, or use an older version of BBTools.
        at dna.Data.setGenome2(Data.java:854)
        at dna.Data.setGenome(Data.java:769)
        at align2.BBMap.loadIndex(BBMap.java:317)
        at align2.BBMap.main(BBMap.java:33)

I'm running it on multiple samples, and it seems to only fail on a handful throwing the error above. Besides, I've run the exact command before with a different reference, using the same version of bbmap, and didn't have any problems then...the only difference being the number of samples (20 in the successful run and 46 in the error-prone run).

Any suggestions are welcome!

Thanks!

Answer 1

[ATpoint]

The index format has changed in this version of BBTools. Please delete the /ref/ directory and re-index from the reference fasta.

Please read the error messages. It is clear what you have to do => re-index the genome if you want to stick with that version of the tool or use an older version of BBTools or go back to the version you used before if you want to stick with the old index.

[drikaul]

Thanks for your response! I saw that, but I'm just confused as to why I would need to re-index my reference fasta each time I run the command on a new sample? Would the error disappear if I first indexed the reference and then ran the command on the pool of samples without the "ref=" argument?

[ATpoint]

The error says that you are right now using a new version of bbmap but the index was build with an older one. YOu only have to index the reference once, but with that version of the tool.

[drikaul]

but I checked the bbmap version for both the runs, and it's the same, since we haven't updated our install of bbmap for a few months now.