Thank you for this great pipeline.
I'm running MetaPro v4.0.3 in an Apptainer container on an HPC (PBS scheduler) using paired-end Illumina metatranscriptomic data.
Metapro fails during the GA_BT2 stage because the generated jobs invoke Bowtie2 against reference databases that only contain BWA indexes.
The downloaded database contains:
family_group/
genus_group/
order_group/
choco_h3_family/
choco_h3_genus/
choco_h3_order/
The family_group directory contains FASTA files together with BWA index files:
XXXX.fasta
XXXX.fasta.amb
XXXX.fasta.ann
XXXX.fasta.bwt
XXXX.fasta.pac
XXXX.fasta.sa
The archived choco_h3_family.tar.gz also seems to contain BWA index files rather than Bowtie2 indexes.
Relevant settings within my config.ini:
source_taxa_db = /scratch/.../metapro_db/family_group
DNA_DB = /scratch/.../metapro_db/family_group
taxid_tree = /scratch/.../metapro_db/taxid_trees/family_tree.tsv #downloaded separately
kraken2_db = /scratch/.../metapro_db
The GA pre-scan completes successfully and produces a populated lib_list.txt.
The generated GA_BT2 job scripts seems to invoke Bowtie2:
bowtie2 \
-x /scratch/.../metapro_db/family_group/2303507.fasta \
-U .../s.fastq \
| samtools view >> .../s.sam
However, that reference only has BWA indexes:
2303507.fasta
2303507.fasta.amb
2303507.fasta.ann
2303507.fasta.bwt
2303507.fasta.pac
2303507.fasta.sa
and therefore Bowtie2 exits with:
(ERR): ".../family_group/2303507.fasta" does not exist or is not a Bowtie 2 index
Exiting now ...
This occurs repeatedly for many taxon-specific databases.
Is this an incompatibility between the v4.0.3 container and the current downloadable database?
Specifically, should the family_group database now contain Bowtie2 (*.bt2) indexes instead of BWA indexes?
Or should the generated GA_BT2 jobs still be using BWA?
thank you for you time
Thank you for this great pipeline.
I'm running MetaPro v4.0.3 in an Apptainer container on an HPC (PBS scheduler) using paired-end Illumina metatranscriptomic data.
Metapro fails during the GA_BT2 stage because the generated jobs invoke Bowtie2 against reference databases that only contain BWA indexes.
The downloaded database contains:
family_group/
genus_group/
order_group/
choco_h3_family/
choco_h3_genus/
choco_h3_order/
The family_group directory contains FASTA files together with BWA index files:
XXXX.fasta
XXXX.fasta.amb
XXXX.fasta.ann
XXXX.fasta.bwt
XXXX.fasta.pac
XXXX.fasta.sa
The archived choco_h3_family.tar.gz also seems to contain BWA index files rather than Bowtie2 indexes.
Relevant settings within my config.ini:
source_taxa_db = /scratch/.../metapro_db/family_group
DNA_DB = /scratch/.../metapro_db/family_group
taxid_tree = /scratch/.../metapro_db/taxid_trees/family_tree.tsv #downloaded separately
kraken2_db = /scratch/.../metapro_db
The GA pre-scan completes successfully and produces a populated lib_list.txt.
The generated GA_BT2 job scripts seems to invoke Bowtie2:
However, that reference only has BWA indexes:
2303507.fasta
2303507.fasta.amb
2303507.fasta.ann
2303507.fasta.bwt
2303507.fasta.pac
2303507.fasta.sa
and therefore Bowtie2 exits with:
This occurs repeatedly for many taxon-specific databases.
Is this an incompatibility between the v4.0.3 container and the current downloadable database?
Specifically, should the family_group database now contain Bowtie2 (*.bt2) indexes instead of BWA indexes?
Or should the generated GA_BT2 jobs still be using BWA?
thank you for you time