initial commit

README.md

@@ -0,0 +1,18 @@
+ # vHRR: co-expression based gene function prediction
+
+ ## Installation
+
+ 1. Clone repo
+ 2. Add path to go.obo file in go_config.sh
+ 3. Run `bash go_config.sh`
+
+ ## Run
+
+ The folders `input` and `nextflow_config` currently contain example data. Refer to this to check data formats.
+
+ 1. Replace the example data `./input/` with your own.
+ 2. Run `bash preprocess_go_data.sh` (replace `go_gene.tsv` with the filename of your input GO annotation file).
+ 3. Replace the input files in `nextflow_config/vHRR_single.config` or `nextflow_config/vHRR_multiple.config` with your filenames and adapt the nextflow parameters to  fit your needs and system.
+ 4. For a single atlas: run `bash vHRR_single.sh`. For multiple atlases: run `bash vHRR_multiple.sh`. 
+
+

go_config.sh

@@ -0,0 +1,3 @@
+module load python
+
+python3 gba_config.py -obo </path/to/go.obo>

nextflow_config/vHRR_multiple.config

@@ -0,0 +1,31 @@
+ // Input data
+params.wkdir = "/group/transreg/thpuy/vHRR"
+params.atlases = "${params.wkdir}/input/example_atlas_*.tsv"
+params.go_gene = "${params.wkdir}/input/go_gene_ic.tsv"
+params.out_folder = "${params.wkdir}/output"
+
+
+executor{
+  queueSize=100
+}
+
+process {
+
+  withName: prepare_starter_pack {
+    executor = 'sge'
+    cpus = 1
+    clusterOptions = '-l h_vmem=1G'
+  }
+
+  withName: scan_HRR {
+    executor = 'sge'
+    cpus = 1
+    clusterOptions = '-l h_vmem=1G'
+  }
+
+  withName: evaluate_predictions {
+    executor = 'sge'
+    cpus = 1
+    clusterOptions = '-l h_vmem=1G'
+  }
+}

nextflow_config/vHRR_single.config

@@ -0,0 +1,32 @@
+ // Input data
+params.wkdir = "/group/transreg/thpuy/vHRR"
+params.atlas = "${params.wkdir}/input/example_atlas_1.tsv"
+params.go_gene = "${params.wkdir}/input/go_gene_ic.tsv"
+params.out_folder = "${params.wkdir}/output"
+params.sample_weighing = ""
+
+
+executor{
+  queueSize=100
+}
+
+process {
+
+  withName: prepare_starter_pack {
+    executor = 'sge'
+    cpus = 1
+    clusterOptions = '-l h_vmem=1G'
+  }
+
+  withName: scan_HRR {
+    executor = 'sge'
+    cpus = 1
+    clusterOptions = '-l h_vmem=1G'
+  }
+
+  withName: evaluate_predictions {
+    executor = 'sge'
+    cpus = 1
+    clusterOptions = '-l h_vmem=1G'
+  }
+}

preprocess_go_data.sh

@@ -0,0 +1,6 @@
+module load python
+
+INFILE=input/go_gene.tsv
+OUTFILE=input/go_gene_ic.tsv
+
+python3 scripts/preprocess_go_data.py $INFILE $OUTFILE

scripts/backend.py

@@ -0,0 +1,14 @@
+import configparser
+from pathlib import Path
+
+def read_config():
+    config = configparser.ConfigParser()
+    try:
+        config.read('{}/.gbaconfig'.format(Path.home()))
+    except:
+        raise FileNotFoundError('[ERROR] GBA config file missing: run scripts/gba_config.py (usage: python3 gba_config.py [options])')
+    return config
+
+def get_obo():
+    config = read_config()
+    return config['GO']['go_obo']

scripts/calculate_ICu_per_GO.py

@@ -0,0 +1,23 @@
+#!/usr/bin/python
+from sys import argv
+from math import log
+
+go_files = argv[1:]
+go_data = dict() # { go_id: set(gene_id) }
+total_genes = set()
+def read_go_file(go_file):
+	with open(go_file, 'r') as reader:
+		for line in reader:
+			go_id, gene_id = line.strip().split("\t")[:2]
+			if not go_id in go_data:
+				go_data[go_id] = set()
+			go_data[go_id].add(gene_id)
+			total_genes.add(gene_id)
+for go_file in go_files:
+	read_go_file(go_file)
+
+total_genes = float(len(total_genes))
+for go_id in sorted(go_data):
+	ICu = -1 * log(float(len(go_data[go_id])) / total_genes, 10) / log(total_genes, 10)
+	print(go_id+"\t"+str(ICu))
+

scripts/extend_go.py

@@ -0,0 +1,14 @@
+#!/usr/bin/python
+import go_manipulations
+from os import path
+from sys import argv
+
+gene_go_file = argv[1]
+go_obo = argv[2]
+
+go_tree = go_manipulations.GOtree(go_obo)
+
+gene_go = go_manipulations.GOgenes(gene_go_file, go_tree)
+go_tree.extend(gene_go)
+gene_go.write()
+

scripts/filter_go_category.py

@@ -0,0 +1,14 @@
+#!/usr/bin/python
+import go_manipulations
+from os import path
+from sys import argv
+
+gene_go_file = argv[1]
+filter_category = argv[2]
+go_obo = argv[3]
+
+go_tree = go_manipulations.GOtree(go_obo)
+
+go_genes = go_manipulations.GOgenes(gene_go_file, go_tree)
+go_tree.filter_category(go_genes, filter_category)
+go_genes.write()

scripts/gba_config.py

@@ -0,0 +1,50 @@
+
+'''
+Setup user to run GBA framework
+For now: only sets the go.obo file
+
+usage: python3 gba_config.py -obo <go.obo>
+'''
+
+
+import argparse
+import configparser
+
+from sys import stderr
+from pathlib import Path
+from goatools.obo_parser import GODag
+
+
+#command line arguments
+parser = argparse.ArgumentParser()
+parser.add_argument("-obo", dest="go_obo", help="set location of go.obo file")
+args = parser.parse_args()
+
+
+# get version info
+obodag_version = GODag(args.go_obo).version.split('rel(')[1].split(')')[0]
+
+
+# init
+config = configparser.ConfigParser()
+
+# set values
+config['GO'] = {'go_obo': args.go_obo}
+
+# write file
+with open('{}/.gbaconfig'.format(Path.home()), 'w') as outf:
+
+    outf.write(
+"""# GBA config file
+# ---------------
+#
+# go.obo : 
+#    - file: {}
+#    - date: {}
+
+""".format(args.go_obo, obodag_version)
+              )
+
+    config.write(outf)
+
+stderr.write('[INFO] GBA config file writen to ~/.gbaconfig\n')