Dev 0.7.6 - Notebook updates

.tests/unit/arts_allhits_combine/data/config/config.yaml

@@ -0,0 +1,64 @@
+# This file should contain everything to configure the workflow on a global scale.
+
+#### PROJECT INFORMATION ####
+# This section control your project configuration.
+# Each project are separated by "-".
+# A project can be defined as (1) a yaml object or (2) a Portable Encapsulated Project (PEP) file.
+# (1) To define project as a yaml object, it must contain the variable "name" and "samples".
+#   - name : name of your project
+#   - samples : a csv file containing a list of genome ids for analysis with multiple sources mentioned. Genome ids must be unique.
+#   - rules: a yaml file containing project rule configurations. This will override global rule configuration.
+#   - prokka-db (optional): list of the custom accessions to use as prokka reference database.
+#   - gtdb-tax (optional): output summary file of GTDB-tk with "user_genome" and "classification" as the two minimum columns
+# (2) To define project using PEP file, only variable "name" should be given that points to the location of the PEP yaml file.
+#   - pep: path to PEP .yaml file. See project example_pep for details.
+# PS: the variable pep and name is an alias
+
+projects:
+# Project 1 (yaml object)
+  - name: config/lactobacillus_delbruecki/project_config.yaml
+
+bgc_projects:
+  - pep: config/lanthipeptide/project_config.yaml
+
+#### GLOBAL RULE CONFIGURATION ####
+# This section configures the rules to run globally.
+# Use project specific rule configurations if you want to run different rules for each projects.
+# pipelines or rules: set value to TRUE if you want to run the analysis or FALSE if you don't
+pipelines:
+  seqfu: FALSE
+  mash: FALSE
+  fastani: FALSE
+  checkm: FALSE
+  gtdbtk: FALSE
+  prokka-gbk: FALSE
+  antismash: TRUE
+  query-bigslice: FALSE
+  bigscape: FALSE
+  bigslice: FALSE
+  automlst-wrapper: FALSE
+  arts: FALSE
+  roary: FALSE
+  eggnog: FALSE
+  eggnog-roary: FALSE
+  deeptfactor: FALSE
+  deeptfactor-roary: FALSE
+  cblaster-genome: FALSE
+  cblaster-bgc: FALSE
+
+#### RESOURCES CONFIGURATION ####
+# resources : the location of the resources to run the rule.
+# The default location is at "resources/{resource_name}".
+resources_path:
+  antismash_db: resources/antismash_db
+  eggnog_db: resources/eggnog_db
+  BiG-SCAPE: resources/BiG-SCAPE
+  bigslice: resources/bigslice
+  checkm: resources/checkm
+  gtdbtk: resources/gtdbtk
+  #RNAmmer: resources/RNAmmer # If specified, will override Barnapp in Prokka
+
+rule_parameters:
+  install_gtdbtk:
+    release: 214
+    release_version: 214

.tests/unit/arts_allhits_combine/data/config/lactobacillus_delbruecki/project_config.yaml

@@ -0,0 +1,30 @@
+name: Lactobacillus_delbrueckii
+
+pep_version: 2.1.0
+
+description: "Lactobacillus delbrueckii 27 01 2023"
+
+sample_table: samples.csv
+
+#### RULE CONFIGURATION ####
+# rules: set value to TRUE if you want to run the analysis or FALSE if you don't
+rules:
+  seqfu: TRUE
+  mash: TRUE
+  fastani: TRUE
+  checkm: FALSE
+  gtdbtk: FALSE
+  prokka-gbk: TRUE
+  antismash: TRUE
+  query-bigslice: TRUE
+  bigscape: TRUE
+  bigslice: TRUE
+  automlst-wrapper: TRUE
+  arts: TRUE
+  roary: TRUE
+  eggnog: TRUE
+  eggnog-roary: TRUE
+  deeptfactor: TRUE
+  deeptfactor-roary: TRUE
+  cblaster-genome: TRUE
+  cblaster-bgc: TRUE

.tests/unit/arts_allhits_combine/data/config/lactobacillus_delbruecki/samples.csv

@@ -0,0 +1,5 @@
+genome_id,source,organism,genus,species,strain,closest_placement_reference,input_file
+GCA_000056065.1,ncbi,,,,,,
+GCA_000182835.1,ncbi,,,,,,
+GCA_000191165.1,ncbi,,,,,,
+GCA_000014405.1,ncbi,,,,,,

.tests/unit/arts_allhits_combine/data/data/interim/bgcs/Lactobacillus_delbrueckii/7.0.0/GCA_000014405.1/GCA_000014405.1-change_log.json

@@ -0,0 +1,22 @@
+{
+    "GCA_000014405.1": {
+        "CP000412.1.region001.gbk": {
+            "record_id": "CP000412.1",
+            "original_id": "CP000412.1",
+            "target_path": "data/interim/antismash/7.0.0/GCA_000014405.1/CP000412.1.region001.gbk",
+            "symlink_path": "data/interim/bgcs/Lactobacillus_delbrueckii/7.0.0/GCA_000014405.1/CP000412.1.region001.gbk"
+        },
+        "CP000412.1.region002.gbk": {
+            "record_id": "CP000412.1",
+            "original_id": "CP000412.1",
+            "target_path": "data/interim/antismash/7.0.0/GCA_000014405.1/CP000412.1.region002.gbk",
+            "symlink_path": "data/interim/bgcs/Lactobacillus_delbrueckii/7.0.0/GCA_000014405.1/CP000412.1.region002.gbk"
+        },
+        "GCA_000014405.1.gbk": {
+            "record_id": "CP000412.1",
+            "original_id": "CP000412.1",
+            "target_path": "data/interim/antismash/7.0.0/GCA_000014405.1/GCA_000014405.1.gbk",
+            "symlink_path": "data/interim/bgcs/Lactobacillus_delbrueckii/7.0.0/GCA_000014405.1/GCA_000014405.1.gbk"
+        }
+    }
+}

.tests/unit/arts_allhits_combine/data/data/interim/bgcs/Lactobacillus_delbrueckii/7.0.0/GCA_000056065.1/GCA_000056065.1-change_log.json

@@ -0,0 +1,28 @@
+{
+    "GCA_000056065.1": {
+        "CR954253.1.region001.gbk": {
+            "record_id": "CR954253.1",
+            "original_id": "CR954253.1",
+            "target_path": "data/interim/antismash/7.0.0/GCA_000056065.1/CR954253.1.region001.gbk",
+            "symlink_path": "data/interim/bgcs/Lactobacillus_delbrueckii/7.0.0/GCA_000056065.1/CR954253.1.region001.gbk"
+        },
+        "CR954253.1.region002.gbk": {
+            "record_id": "CR954253.1",
+            "original_id": "CR954253.1",
+            "target_path": "data/interim/antismash/7.0.0/GCA_000056065.1/CR954253.1.region002.gbk",
+            "symlink_path": "data/interim/bgcs/Lactobacillus_delbrueckii/7.0.0/GCA_000056065.1/CR954253.1.region002.gbk"
+        },
+        "CR954253.1.region003.gbk": {
+            "record_id": "CR954253.1",
+            "original_id": "CR954253.1",
+            "target_path": "data/interim/antismash/7.0.0/GCA_000056065.1/CR954253.1.region003.gbk",
+            "symlink_path": "data/interim/bgcs/Lactobacillus_delbrueckii/7.0.0/GCA_000056065.1/CR954253.1.region003.gbk"
+        },
+        "GCA_000056065.1.gbk": {
+            "record_id": "CR954253.1",
+            "original_id": "CR954253.1",
+            "target_path": "data/interim/antismash/7.0.0/GCA_000056065.1/GCA_000056065.1.gbk",
+            "symlink_path": "data/interim/bgcs/Lactobacillus_delbrueckii/7.0.0/GCA_000056065.1/GCA_000056065.1.gbk"
+        }
+    }
+}

.tests/unit/arts_allhits_combine/data/data/interim/bgcs/Lactobacillus_delbrueckii/7.0.0/GCA_000182835.1/GCA_000182835.1-change_log.json

@@ -0,0 +1,10 @@
+{
+    "GCA_000182835.1": {
+        "GCA_000182835.1.gbk": {
+            "record_id": "CP002342.1",
+            "original_id": "CP002342.1",
+            "target_path": "data/interim/antismash/7.0.0/GCA_000182835.1/GCA_000182835.1.gbk",
+            "symlink_path": "data/interim/bgcs/Lactobacillus_delbrueckii/7.0.0/GCA_000182835.1/GCA_000182835.1.gbk"
+        }
+    }
+}

.tests/unit/arts_allhits_combine/data/data/interim/bgcs/Lactobacillus_delbrueckii/7.0.0/GCA_000191165.1/GCA_000191165.1-change_log.json

@@ -0,0 +1,22 @@
+{
+    "GCA_000191165.1": {
+        "CP000156.1.region001.gbk": {
+            "record_id": "CP000156.1",
+            "original_id": "CP000156.1",
+            "target_path": "data/interim/antismash/7.0.0/GCA_000191165.1/CP000156.1.region001.gbk",
+            "symlink_path": "data/interim/bgcs/Lactobacillus_delbrueckii/7.0.0/GCA_000191165.1/CP000156.1.region001.gbk"
+        },
+        "CP000156.1.region002.gbk": {
+            "record_id": "CP000156.1",
+            "original_id": "CP000156.1",
+            "target_path": "data/interim/antismash/7.0.0/GCA_000191165.1/CP000156.1.region002.gbk",
+            "symlink_path": "data/interim/bgcs/Lactobacillus_delbrueckii/7.0.0/GCA_000191165.1/CP000156.1.region002.gbk"
+        },
+        "GCA_000191165.1.gbk": {
+            "record_id": "CP000156.1",
+            "original_id": "CP000156.1",
+            "target_path": "data/interim/antismash/7.0.0/GCA_000191165.1/GCA_000191165.1.gbk",
+            "symlink_path": "data/interim/bgcs/Lactobacillus_delbrueckii/7.0.0/GCA_000191165.1/GCA_000191165.1.gbk"
+        }
+    }
+}

.tests/unit/arts_allhits_combine/data/workflow

@@ -0,0 +1 @@
+../../../../workflow