Dev 0.7.5

workflow/rules/antismash.smk

@@ -69,20 +69,35 @@ elif antismash_major_version >= 7:
             genefinding="none",
         shell:
             """
-            # Find the latest existing json output for this strain
-            latest_version=$(ls -d data/interim/antismash/*/*/*.json | grep {wildcards.strains} | sort -r | head -n 1 | cut -d '/' -f 4)
+            set +e
+
+            # Find the latest existing JSON output for this strain
+            latest_version=$(ls -d data/interim/antismash/*/*/*.json | grep {wildcards.strains} | sort -r | head -n 1 | cut -d '/' -f 4) 2>> {log}
+
             if [ -n "$latest_version" ]; then
-                OLD_JSON="data/interim/antismash/$latest_version/{wildcards.strains}/{wildcards.strains}.json"
-                echo "Using existing json from $OLD_JSON as starting point..." >> {log}
-                ANTISMASH_INPUT="--reuse-result $OLD_JSON"
+                # Use existing JSON result as starting point
+                old_json="data/interim/antismash/$latest_version/{wildcards.strains}/{wildcards.strains}.json"
+                echo "Using existing JSON from $old_json as starting point..." >> {log}
+                antismash_input="--reuse-result $old_json"
             else
-                echo "No existing output directories found, starting AntiSMASH from scratch..." >> {log}
-                ANTISMASH_INPUT="{input.gbk}"
+                # No existing JSON result found, use genbank input
+                echo "No existing JSON result found, starting AntiSMASH from scratch..." >> {log}
+                antismash_input="{input.gbk}"
             fi
-            # run antismash
+
+            # Run AntiSMASH
             antismash --genefinding-tool {params.genefinding} --output-dir {params.folder} \
                 --database {input.resources} \
-                --cb-general --cb-subclusters --cb-knownclusters -c {threads} $ANTISMASH_INPUT --logfile {log} 2>> {log}
+                --cb-general --cb-subclusters --cb-knownclusters -c {threads} $antismash_input --logfile {log} 2>> {log}
+
+            # Check if the run failed due to changed detection results
+            if grep -q "ValueError: Detection results have changed. No results can be reused" {log}; then
+                # Use genbank input instead
+                echo "Previous JSON result is invalid, starting AntiSMASH from scratch..." >> {log}
+                antismash --genefinding-tool {params.genefinding} --output-dir {params.folder} \
+                    --database {input.resources} \
+                    --cb-general --cb-subclusters --cb-knownclusters -c {threads} {input.gbk} --logfile {log} 2>> {log}
+            fi
             """
 
 rule copy_antismash:

workflow/rules/common.smk

@@ -8,7 +8,7 @@ from pathlib import Path
 import peppy
 
 min_version("7.14.0")
-__version__ = "0.7.4"
+__version__ = "0.7.5"
 
 
 container: "docker://matinnu/bgcflow:latest"

workflow/rules/deeptfactor.smk

@@ -14,30 +14,30 @@ rule deeptfactor_setup:
 
 rule deeptfactor:
     input:
-        fasta="data/interim/prokka/{strains}/{strains}.faa",
+        faa="data/interim/prokka/{strains}/{strains}.faa",
         resource="resources/deeptfactor/",
     output:
-        deeptfactor_dir=directory("data/interim/deeptfactor/{strains}/"),
+        deeptfactor="data/interim/deeptfactor/{strains}/prediction_result.txt",
     conda:
         "../envs/deeptfactor.yaml"
     threads: 2
     params:
-        faa="../../data/interim/prokka/{strains}/{strains}.faa",
-        outdir="../../data/interim/deeptfactor/{strains}/",
+        outdir="data/interim/deeptfactor/{strains}/",
     log:
         "logs/deeptfactor/deeptfactor/deeptfactor-{strains}.log",
     shell:
         """
+        workdir=$PWD
         mkdir -p data/interim/deeptfactor/{wildcards.strains} 2>> {log}
         (cd {input.resource} && python tf_running.py \
-            -i {params.faa} -o {params.outdir} \
+            -i $workdir/{input.faa} -o $workdir/{params.outdir} \
             -g cpu -cpu {threads}) 2>> {log}
         """
 
 
 rule deeptfactor_to_json:
     input:
-        deeptfactor_dir="data/interim/deeptfactor/{strains}/",
+        deeptfactor="data/interim/deeptfactor/{strains}/prediction_result.txt",
     output:
         deeptfactor_json="data/interim/deeptfactor/{strains}_deeptfactor.json",
     conda:
@@ -46,7 +46,7 @@ rule deeptfactor_to_json:
         "logs/deeptfactor/deeptfactor/deeptfactor-{strains}_to_json.log",
     shell:
         """
-        python workflow/bgcflow/bgcflow/data/deeptfactor_scatter.py {input.deeptfactor_dir}/prediction_result.txt {output.deeptfactor_json} 2>> {log}
+        python workflow/bgcflow/bgcflow/data/deeptfactor_scatter.py {input.deeptfactor} {output.deeptfactor_json} 2>> {log}
         """
 
 

workflow/rules/roary.smk

@@ -7,8 +7,8 @@ rule roary:
         "../envs/roary.yaml"
     params:
         i=80,
-        g=60000,
-    threads: 8
+        g=80000,
+    threads: 16
     log:
         "logs/roary/roary-{name}.log",
     shell: