Closed MostafaYA closed 1 year ago
The --ignore
option seems to solve the problem!
I close this!!
Hi @MostafaYA ,
No problems mate - thankfully this was suggested by Alex Weisberg.
A bit of a self plug, but I'd also give Hybracter https://github.com/gbouras13/hybracter a try as it should handle these cases for you automatically (I'd love some feedback!) :)
I also know Dnaapler will probably soon be in Dragonflye https://github.com/rpetit3/dragonflye - so that would be another good option to try. Just tagging you in here too @rpetit3 I'm not sure how you have handled non-circularised contigs but this concern is something to think about if you haven't thought of it!
George
Hi, thanks for the great pipelines. Below is my code to auto-ignore the non-circular contigs
rule noncircular_contigs:
input:
rules.copy_flye_data.output[1],
output:
results_per_sample_dir + "{sample}/4_genome_rotation/noncircular_contigs.tsv",
conda:
envs_folder + "csvtk.yaml"
shell:
"""
csvtk filter2 -t -C $ -f '$4== "N" ' {input} -o {output}
"""
rule rotate_genome:
input:
rotate_genome_input(),
rules.noncircular_contigs.output[0],
output:
dir = temp(directory(results_per_sample_dir + "{sample}/tmp_rotation")),
contigs= results_per_sample_dir + "{sample}/4_genome_rotation/{sample}.fasta",
summary= results_per_sample_dir + "{sample}/4_genome_rotation/{sample}.reorientation_summary.tsv"
threads: 16
conda:
envs_folder + "dnaapler.yaml"
log: results_per_sample_dir + "{sample}/4_genome_rotation/dnaapler.log"
params:
options = "--force "
shell:
"""
dnaapler all {params.options} \
--prefix {wildcards.sample} \
--output {output.dir} \
--threads {threads} \
--ignore {input[1]} \
--input {input[0]} 2>&1 | tee {log} >/dev/null
cp {output.dir}/{wildcards.sample}_reoriented.fasta {output.contigs}
cp {output.dir}/{wildcards.sample}_all_reorientation_summary.tsv {output.summary}
"""
Hi,
Thanks for this nice tool. I tried to automate
Dnaapler
in a pipeline to reorientate genomes assembled withflye
. In some instances, I got the genome assembly fragmented, and obviously not circular (also from flye info). Despite of this,Dnaapler
rotated these fragments and may have falsely merged the contig ends. Could it be possible thatDnaapler
takes the flyeassembly_info.txt
and applies its logic only to the circularized contigs?Thank you