changes for version 1.3.1

config.py.sample

@@ -0,0 +1,165 @@
+#!/usr/bin/env python
+
+import getpass
+import os
+
+# Required user input:
+# 1) Which fusion prediction tools should be executed (tools)
+# 2) Which post-processing steps should be executed (fd_tools)
+# 3) Which reference data shall be used (ref_trans_version & ref_genome_build)
+# 4) To whom shall slurm mails be sent to (receiver)
+
+__version__ = "1.3.1"
+
+pipeline_name = "EasyFuse"
+
+#tools=Readfilter,Mapsplice,Fusioncatcher,Star,Starfusion,Infusion,Fetchdata,Summary
+tools = ("QC", 
+         "Readfilter", 
+         "Fusioncatcher", 
+         "Star", 
+         "Starfusion", 
+         "Infusion", 
+         "Mapsplice", 
+         "Soapfuse", 
+         "Fetchdata", 
+         "Summary")
+
+fusiontools = ("Fusioncatcher", 
+               "Starfusion", 
+               "Infusion", 
+               "Mapsplice", 
+               "Soapfuse")
+
+#fd_tools=Fusiongrep,Contextseq,Starindex,Staralign,Bamindex,Requantify
+fd_tools = ("Fusiongrep", 
+            "Contextseq", 
+            "Starindex", 
+            "ReadFilter2", 
+            "ReadFilter2b", 
+            "StaralignBest", 
+            "BamindexBest", 
+            "RequantifyBest")
+
+sender = "[email protected]"
+receiver = "[email protected]"
+min_read_len_perc = 0.75
+max_dist_proper_pair = 200000
+cis_near_distance = 1000000
+model_pred_threshold = 0.75
+tsl_filter = "4,5,NA"
+requant_mode = ["best"]
+context_seq_len = 400
+ref_genome_build = "hg38"
+ref_trans_version = "ensembl"
+queueing_system = "slurm"
+time_limit = "30-00:00:0"
+partition = "Compute"
+user = getpass.getuser()
+module_dir = os.path.dirname(os.path.realpath(__file__))
+#logfile=/data/urla_progs/TronData/ngs_pipelines/easyfuse/fusion.log
+#fusion_db=/data/urla_progs/TronData/ngs_pipelines/easyfuse/fusion.db
+
+# Define ressource usage (cpu (number of threads), mem (ram in Gb)):
+resources = {
+    "qc": {
+        "cpu": 6,
+        "mem": 10
+    },
+    "readfilter": {
+        "cpu": 6,
+        "mem": 50
+    },
+    "star": {
+        "cpu": 6,
+        "mem": 40
+    },
+    "kallisto": {
+        "cpu": 6,
+        "mem": 10
+    },
+    "mapsplice": {
+        "cpu": 6,
+        "mem": 30
+    },
+    "fusioncatcher": {
+        "cpu": 6,
+        "mem": 30
+    },
+    "starfusion": {
+        "cpu": 6,
+        "mem": 30
+    },
+    "starchip": {
+        "cpu": 12,
+        "mem": 30
+    },
+    "infusion": {
+        "cpu": 6,
+        "mem": 30
+    },
+    "soapfuse": {
+        "cpu": 6,
+        "mem": 20
+    },
+    "classification": {
+        "cpu": 1,
+        "mem": 16
+    },
+    "fetchdata": {
+        "cpu": 12,
+        "mem": 50
+    },
+    "summary": {
+        "cpu": 1,
+        "mem": 16
+    }
+}
+
+# execution command for individual programs (what you write here should be identical to what is typed in the console)
+commands = {
+    # for qc
+    "fastqc": "/path/to/fastqc_bin",
+    "skewer": "/path/to/skewer_bin",
+    # for processing
+    "mapsplice": "/path/to/mapsplice.py",
+    "fusioncatcher": "/path/to/fusioncatcher_bin",
+    "starfusion": "/path/to/starfusion_bin",
+    "infusion": "/path/to/infusion_bin",
+    "soapfuse": "/path/to/SOAPfuse-RUN.pl",
+    # for processing and fetch data
+    "star": "/path/to/star_bin",
+    "samtools": "/path/to/samtools_bin"
+    # for liftover
+    #"crossmap": "test"
+}
+
+# full path to reference files
+references = {
+    # wget http://ftp.ensembl.org/pub/release-86/fasta/homo_sapiens/dna/Homo_sapiens.GRCh38.dna.primary_assembly.fa.gz
+    "genome_fasta": "/path/to/Homo_sapiens.GRCh38.dna.primary_assembly.fa",
+    "genome_fastadir": "/projects/data/human/ensembl/GRCh38.86/fasta",
+    "genome_sizes": "/projects/data/human/ensembl/GRCh38.86/STAR_idx/chrNameLength.txt",
+    # wget http://ftp.ensembl.org/pub/release-86/fasta/homo_sapiens/cdna/Homo_sapiens.GRCh38.cdna.all.fa.gz
+    "genes_fasta": "/projects/data/human/ensembl/GRCh38.86/Homo_sapiens.GRCh38.cdna.all.fa",
+    # wget http://ftp.ensembl.org/pub/release-86/gtf/homo_sapiens/Homo_sapiens.GRCh38.86.chr.gtf.gz
+    "genes_gtf": "/path/to/Homo_sapiens.GRCh38.86.gtf",
+    "genes_adb": "/projects/data/human/ensembl/GRCh38.86/Homo_sapiens.GRCh38.86.gff3.db",
+    "genes_tsl": "/projects/data/human/ensembl/GRCh38.86/Homo_sapiens.GRCh38.86.gtf.tsl"
+}
+
+# full path to program indices
+indices = {
+    "star": "/projects/data/human/ensembl/GRCh38.86/STAR_idx/",
+    "bowtie": "/projects/data/human/ensembl/GRCh38.86/bowtie_index/hg38",
+    "starfusion": "/projects/data/human/ensembl/GRCh38.86/starfusion_index/",
+    "fusioncatcher": "/projects/data/human/ensembl/GRCh38.86/fusioncatcher_index/"
+}
+
+# full path to program specific config files (these are just general files which need no user modification)
+other_files = {
+    "infusion_cfg": "/projects/data/human/ensembl/GRCh38.86/infusion_index/infusion.cfg",
+    "soapfuse_cfg": "/code/SOAPfuse/1.27/config/config_h86.txt",
+    "soapfuse_cfg_mm10": "/code/SOAPfuse/1.27/config/config_m95.txt",
+    "easyfuse_model": os.path.join(module_dir, "data", "model", "Fusion_modeling_IVAC_BNT_v16.model.requant_and_boundary_org.randomForest.rds")
+}