added activity plot filtering parameter

pinellolab · Dec 12, 2017 · 64ed75f · 64ed75f
1 parent 8626f42
commit 64ed75f
Show file tree

Hide file tree

Showing 7 changed files with 43 additions and 18 deletions.
diff --git a/README.md b/README.md
@@ -3,7 +3,7 @@
 
 Epigenetic Variability and Motif Analysis Pipeline       
 --------------------------------------------------
-**Current version**: 0.5.2
+**Current version**: 0.5.3
 
 [![Build Status](https://travis-ci.org/pinellolab/haystack_bio.svg?branch=master)](https://travis-ci.org/pinellolab/haystack_bio)
 

diff --git a/haystack/download_genome.py b/haystack/download_genome.py
@@ -1,8 +1,8 @@
 
-from haystack_common import initialize_genome
+from haystack_common import initialize_genome, HAYSTACK_VERSION
 import argparse
 import sys
-HAYSTACK_VERSION = "0.5.2"
+
 def get_args_download_genome():
     # mandatory
     parser = argparse.ArgumentParser(description='download_genome parameters')

diff --git a/haystack/find_hotspots.py b/haystack/find_hotspots.py
@@ -8,7 +8,7 @@
 #from pybedtools import BedTool
 import multiprocessing
 import glob
-from haystack_common import determine_path, check_file, check_required_packages, initialize_genome
+from haystack_common import determine_path, check_file, check_required_packages, initialize_genome, HAYSTACK_VERSION
 
 import logging
 logging.basicConfig(level=logging.INFO,
@@ -21,8 +21,6 @@
 debug = logging.debug
 info = logging.info
 
-HAYSTACK_VERSION = "0.5.2"
-
 do_not_recompute = None
 keep_intermediate_files= None
 

diff --git a/haystack/find_motifs.py b/haystack/find_motifs.py
@@ -18,8 +18,8 @@
     import pickle as cp
 
 # commmon functions
-from haystack_common import determine_path, which, check_file, initialize_genome
-HAYSTACK_VERSION = "0.5.2"
+from haystack_common import determine_path, which, check_file, initialize_genome, HAYSTACK_VERSION
+
 # dependencies
 from bioutilities import  Coordinate, Sequence, Fimo
 import numpy as np

diff --git a/haystack/generate_tf_activity_plane.py b/haystack/generate_tf_activity_plane.py
@@ -12,9 +12,8 @@
 import argparse
 from itertools import chain
 import logging
-from haystack_common import determine_path, check_file
+from haystack_common import determine_path, check_file, HAYSTACK_VERSION
 
-HAYSTACK_VERSION = "0.5.2"
 
 logging.basicConfig(level=logging.INFO,
                     format='%(levelname)-5s @ %(asctime)s:\n\t %(message)s \n',
@@ -77,6 +76,14 @@ def get_args_activity():
     parser.add_argument('--plot_all',
                         help='Disable the filter on the TF activity and correlation (default z-score TF>0 and rho>0.3)',
                         action='store_true')
+    parser.add_argument('--rho_cutoff',
+                        type=float,
+                        default=0.3,
+                        help='The cutoff absolute correlation value (0.0 to 1)  for which activity plots are generated (default: 0.3)')
+    parser.add_argument('--tf_value_cuttoff',
+                        type=float,
+                        default=0.0,
+                        help='The cutoff z-score tf_value for which activity plots are generated (default: 0.0) ')
     parser.add_argument('--version',
                         help='Print version and exit.',
                         action='version',
@@ -209,16 +216,16 @@ def main(input_args=None):
 
                 if USE_ZSCORE:
                     # correlation
-                    ro = np.corrcoef(tf_values, ds_values)[0, 1]
+                    rho = np.corrcoef(tf_values, ds_values)[0, 1]
 
                     tf_value = tf_values[target_cell_type]
                     ds_value = ds_values[target_cell_type]
 
                     info('Gene:%s TF z-score:%.2f Targets z-score:%.2f  Correlation:%.2f' % (
-                        gene_name, tf_value, ds_value, ro))
+                        gene_name, tf_value, ds_value, rho))
 
                     # make plots
-                    if (tf_value > 0 and np.abs(ro) > 0.3) or args.plot_all:
+                    if (tf_value >  args.tf_value_cuttoff and np.abs(rho) > args.rho_cutoff) or args.plot_all:
                         x_min = min(-4, tf_values.min() * 1.1)
                         x_max = max(4, tf_values.max() * 1.1)
                         y_min = min(-4, ds_values.min() * 1.1)
@@ -237,7 +244,7 @@ def main(input_args=None):
                                   numpoints=1)
 
                         ax.set_aspect('equal')
-                        plt.text(x_min * 0.98, y_max * 0.85, r'$\rho$=%.2f' % ro, fontsize=14)
+                        plt.text(x_min * 0.98, y_max * 0.85, r'$\rho$=%.2f' % rho, fontsize=14)
                         plt.xlim(x_min, x_max)
                         plt.ylim(y_min, y_max)
 

diff --git a/haystack/haystack_common.py b/haystack/haystack_common.py
@@ -18,6 +18,10 @@
 debug = logging.debug
 info = logging.info
 
+
+__version__ = "0.5.3"
+HAYSTACK_VERSION=__version__
+
 def check_file(filename):
     try:
         with open(filename): pass

diff --git a/haystack/run_pipeline.py b/haystack/run_pipeline.py
@@ -8,7 +8,7 @@
 
 import subprocess as sb
 
-from haystack_common import check_file
+from haystack_common import check_file, HAYSTACK_VERSION
 
 import logging
 logging.basicConfig(level=logging.INFO,
@@ -25,7 +25,6 @@
 #from memory_profiler import profile
 #f = open('pipeline_memory.txt', 'w+')
 
-HAYSTACK_VERSION = "0.5.2"
 
 def get_args_pipeline():
     # mandatory
@@ -89,6 +88,14 @@ def get_args_pipeline():
                         default='_|chrX|chrY')
     parser.add_argument('--read_ext', type=int, help='Read extension in bps (default: 200)', default=200)
     parser.add_argument('--temp_directory', help='Directory to store temporary files  (default: /tmp)', default='/tmp')
+    parser.add_argument('--rho_cutoff',
+                        type=float,
+                        default=0.3,
+                        help='The cutoff absolute correlation value (0.0 to 1)  for which activity plots are generated (default: 0.3)')
+    parser.add_argument('--tf_value_cuttoff',
+                        type=float,
+                        default=0.0,
+                        help='The cutoff z-score tf_value for which activity plots are generated (default: 0.0) ')
     parser.add_argument('--version', help='Print version and exit.', action='version',
                         version='Version %s' % HAYSTACK_VERSION)
 
@@ -289,16 +296,25 @@ def main(input_args=None):
                                                 'HAYSTACK_MOTIFS_on_%s' % sample_name)
 
             if os.path.exists(motifs_output_folder):
-                cmd_to_run='haystack_tf_activity_plane %s %s %s --output_directory %s'  %(motifs_output_folder,sample_names_tf_activity_filename,sample_name,tf_activity_directory)
+                cmd_to_run='haystack_tf_activity_plane %s %s %s --output_directory %s'  %(motifs_output_folder,
+                                                                                          sample_names_tf_activity_filename,
+                                                                                          sample_name,
+                                                                                          tf_activity_directory)
 
             if motif_mapping_filename:
                 cmd_to_run+=' --motif_mapping_filename %s' %  motif_mapping_filename
 
             if plot_all:
                 cmd_to_run+=' --plot_all'
 
+            if rho_cutoff:
+                cmd_to_run += ' --rho_cutoff %f' % rho_cutoff
+
+            if tf_value_cuttoff:
+                cmd_to_run += ' --tf_value_cuttoff %f' % tf_value_cuttoff
+
             print(cmd_to_run)
-            sb.call(cmd_to_run,shell=True)
+            sb.call(cmd_to_run, shell=True)
 
 
 if __name__ == '__main__':