Skip to content
Snippets Groups Projects
Commit 214f07ad authored by Valentina Galata's avatar Valentina Galata
Browse files

cdhit: changed renaming rule (issue #75), use '-s2 0.5' (issue #68)

parent ae21c8ce
No related branches found
No related tags found
No related merge requests found
...@@ -28,27 +28,29 @@ rule analysis_quast: ...@@ -28,27 +28,29 @@ rule analysis_quast:
# Proteins # Proteins
# for CD-HIT # for CD-HIT
rule analysis_bbmap_rename: rule analysis_cdhit_rename:
input: input:
os.path.join(RESULTS_DIR, "annotation/prodigal/{rtype}/{tool}/proteins.faa") os.path.join(RESULTS_DIR, "annotation/prodigal/{rtype}/{tool}/proteins.faa")
output: output:
temp(os.path.join(RESULTS_DIR, "analysis/cdhit/{rtype}_{tool}.faa")) temp(os.path.join(RESULTS_DIR, "analysis/cdhit/{rtype}_{tool}.faa"))
log: # log:
"logs/analysis_bbmap_rename_{rtype}.{tool}.log" # "logs/analysis_bbmap_rename_{rtype}.{tool}.log"
wildcard_constraints: wildcard_constraints:
rtype="|".join(READ_TYPES), rtype="|".join(READ_TYPES),
tool="|".join(ASSEMBLERS) tool="|".join(ASSEMBLERS)
threads: threads: 1
config["bbmap"]["threads"] # config["bbmap"]["threads"]
params: # params:
# TODO: is this prefix okay? # # TODO: is this prefix okay?
prefix=lambda wildcards: "%s_%s" % (wildcards.rtype, wildcards.tool) # prefix=lambda wildcards: "%s_%s" % (wildcards.rtype, wildcards.tool)
conda: # conda:
os.path.join(ENV_DIR, "bbmap.yaml") # os.path.join(ENV_DIR, "bbmap.yaml")
message: message:
"BBMAP: rename FASTA entries in {input}" # "BBMAP: rename FASTA entries in {input}"
"CDHIT: rename FASTA entries in {input}"
shell: shell:
"(date && rename.sh in={input} out={output} prefix={params.prefix} ignorejunk=t && date) &> {log}" # "(date && rename.sh in={input} out={output} prefix={params.prefix} ignorejunk=t && date) &> {log}"
"sed 's/^>/>{wildcards.rtype}__{wildcards.tool}__/' {input} > {output}"
# https://github.com/weizhongli/cdhit/wiki/3.-User's-Guide#CDHIT2D # https://github.com/weizhongli/cdhit/wiki/3.-User's-Guide#CDHIT2D
rule analysis_cdhit: rule analysis_cdhit:
...@@ -73,8 +75,8 @@ rule analysis_cdhit: ...@@ -73,8 +75,8 @@ rule analysis_cdhit:
"CD-HIT: {input}" "CD-HIT: {input}"
shell: shell:
"(date && " "(date && "
"cd-hit-2d -i {input.faa1} -i2 {input.faa2} -o {output.faa12} -c 0.9 -n 5 -d 0 -M 16000 -s2 0.9 -T {threads} && " "cd-hit-2d -i {input.faa1} -i2 {input.faa2} -o {output.faa12} -c 0.9 -n 5 -d 0 -M 16000 -s2 0.5 -T {threads} && "
"cd-hit-2d -i {input.faa2} -i2 {input.faa1} -o {output.faa21} -c 0.9 -n 5 -d 0 -M 16000 -s2 0.9 -T {threads} && " "cd-hit-2d -i {input.faa2} -i2 {input.faa1} -o {output.faa21} -c 0.9 -n 5 -d 0 -M 16000 -s2 0.5 -T {threads} && "
"date) &> {log}" "date) &> {log}"
################################################## ##################################################
......
0% Loading or .
You are about to add 0 people to the discussion. Proceed with caution.
Finish editing this message first!
Please register or to comment