Hi Daniel, I ran the pipeline with `--user-orfs-f...
# metatdenovo
m
Hi Daniel, I ran the pipeline with
--user-orfs-faa
and
--user-orfs-gff
and mistakenly defined
--orf-caller
. I see in the working directory, the
.command.sh
for the process
NFCORE_METATDENOVO:METATDENOVO:KOFAMSCAN:KOFAMSCAN_SCAN
used the correct input:
Copy code
exec_annotation \
    --profile profiles \
    --ko-list ko_list \
    --format detail-tsv \
    --cpu 6 \
    XXXXX.okay.aa \
    -o kofamscan_output.tsv
but in the results directory the naming is:
Copy code
results/kofamscan/sample_XXX.transdecoder.kofamscan_output.tsv.gz
I guess it is a naming issue?? I mean "transdecoder" has nothing to do with this output! right
d
Did you set
--orf-caller
to
transdecoder
? I'm not sure what would happen in this case, was
transdecoder
invoked?
m
yes as mentioned I forgot to unset
orf-caller
so its value was not
null
but
transdecoder
. I believe in this case, the file names are wrong. Maybe there should be a checkpoint either
orf-caller
or
user-orfs-name
to be used. need a bug? thanks.
d
I thought I had fixed this...
orf_caller
doesn't have a default in the current version (but the docs say differently). Which version did you run?
m
dev
but on a rerun I unset the
orf-caller
and the output files are correctly named.
d
Right, then your output files should have
user_orfs
in their names, unless you set
--user_orfs_name
to something else.
(My reply was before I saw your latest comment.)
👍 1
As I've been into before, I'm fond of our output in the
summary_tables
directory, so this is what I automatically check. I might hence, again, miss naming in the other output directories.