Description of the bug
Ran 10 Shigella flexneri isolates through Phoenix but during the CHECK_SHIGAPASS_TAXA process, the pipeline seems to extract the Shigella species information from the ShigaPass summary file using field 10 (cut -d';' -f10), but I believe the actual species designation (Predicted_FlexSerotype) is in column 8. As a result, the variable is empty, causing downstream species checks between ShigaPass and FastANI to fail.
Example affected file:
PNUSAE091421_ShigaPass_summary.csv:
Name;rfb;rfb_hits,(%);MLST;fliC;CRISPR;ipaH;Predicted_Serotype;Predicted_FlexSerotype;Comments
PNUSAE091421;B1-5;145,(97.3%);none;ShH2(ShH2cplx);A-var3,x,16;ipaH+;SF1-5;2a;
The current process:
sed -n '2p' | cut -d';' -f10
returns an empty string.
The correct process needs to cut -d';' -f8 for Predicted_FlexSerotype, which carries the value (e.g. SF1-5, SS, SB, etc.).
This causes downstream logic to fail, especially when ShigaPass and FastANI assign discordant taxa.
Command used and terminal output
$ nextflow run /Projects/Production/phoenix_v2.3.1 -profile docker --mode PHOENIX --input samplesheet.csv --kraken2db /apps/references/db/k2_standard_08/ --create_ncbi_sheet -resume -c nextflow.config
[...]
cat PNUSAE091421_ShigaPass_summary.csv | sed -n '2p' | cut -d';' -f10
Returns nothing due to the field being empty. Field 8 has the expected serotype value.
Common error:
ValueError: No line with Escherichia_coli found.
Relevant files
.nextflow.log
PNUSAE091421_ShigaPass_summary.csv
modules/local/check_shigapass_taxa.nf
Explanation of how semicolon field counts relate to ShigaPass documentation.
System information
- Nextflow version: 25.04.2
- Container: Docker
- Hardware: Linux VM
- Executor: local
Description of the bug
Ran 10 Shigella flexneri isolates through Phoenix but during the CHECK_SHIGAPASS_TAXA process, the pipeline seems to extract the Shigella species information from the ShigaPass summary file using field 10 (
cut -d';' -f10), but I believe the actual species designation (Predicted_FlexSerotype) is in column 8. As a result, the variable is empty, causing downstream species checks between ShigaPass and FastANI to fail.Example affected file:
PNUSAE091421_ShigaPass_summary.csv:
Name;rfb;rfb_hits,(%);MLST;fliC;CRISPR;ipaH;Predicted_Serotype;Predicted_FlexSerotype;Comments
PNUSAE091421;B1-5;145,(97.3%);none;ShH2(ShH2cplx);A-var3,x,16;ipaH+;SF1-5;2a;
The current process:
sed -n '2p' | cut -d';' -f10
returns an empty string.
The correct process needs to cut -d';' -f8 for Predicted_FlexSerotype, which carries the value (e.g. SF1-5, SS, SB, etc.).
This causes downstream logic to fail, especially when ShigaPass and FastANI assign discordant taxa.
Command used and terminal output
Relevant files
.nextflow.log
PNUSAE091421_ShigaPass_summary.csv
modules/local/check_shigapass_taxa.nf
Explanation of how semicolon field counts relate to ShigaPass documentation.
System information