SPT version 1.0.74 including strand-coordinated mutagenesis colormap is changed, pandas FutureWarning is eliminated, sys.stderr is redirected, warning in TopographyCommons.py is caught

Author: burcakotlu

.gitignore

@@ -2,7 +2,7 @@
 *.pbs
 *.txt
 *.csv
-*.png
+#*.png
 *.wig
 *.bigWig
 *.bed

SigProfilerTopography.png

@@ -2700,6 +2700,7 @@ def runAnalyses(genome,
     print('#################################################################################\n', file=log_out)
     log_out.close()
     sys.stderr.close()
+    sys.stderr = sys.__stderr__ # redirect to original stderr
 
 
 # Plot figures for the attainded data after SigProfilerTopography Analyses

SigProfilerTopography/Topography.py

@@ -31,7 +31,7 @@
 from statsmodels.stats.weightstats import ztest
 
 # To handle warnings as errors
-# import warnings
+import warnings
 # warnings.filterwarnings("error")
 
 LINUX = 'linux'
@@ -654,13 +654,14 @@
 def calculate_pvalue_teststatistics(observed_value,
                                     expected_values,
                                     alternative = 'two-sided'):
-
-    try:
-        zstat, pvalue = ztest(expected_values, [observed_value], alternative=alternative) # pvalue 0.0115612696237375
-    except RuntimeWarning as w:
-        zstat = None
-        pvalue = None
-        # print('expected_values:', expected_values, 'observed_value:', observed_value, 'alternative:', alternative, 'w:', w)
+    with warnings.catch_warnings():
+        warnings.filterwarnings('error')
+        try:
+            zstat, pvalue = ztest(expected_values, [observed_value], alternative=alternative) # pvalue 0.0115612696237375
+        except Warning as w:
+            zstat = None
+            pvalue = None
+            # print('expected_values:', expected_values, 'observed_value:', observed_value, 'alternative:', alternative, 'w:', w)
 
     # zstat, pvalue = ztest(expectedValues, value=observedValue) results in very small p-values therefore we are not calling in this way. # (-25.37854961568692, 4.351195335930552e-142)
     # stats.ttest_1samp(expected_values, observed_value) # Ttest_1sampResult(statistic=-25.378549615686918, pvalue=3.99359102646761e-45)
@@ -3182,7 +3183,7 @@ def readChrBasedMutationsMergeWithProbabilitiesAndWrite(inputList):
         # UAD-US_SP50263 10      110099884       Q:T[GC>AG]C     0
         if simNum >= 1:
             # Get rid of simulation number at the end
-            chr_based_mutation_df[SAMPLE] = chr_based_mutation_df[SAMPLE].str.rsplit('_', 1, expand=True)[0]
+            chr_based_mutation_df[SAMPLE] = chr_based_mutation_df[SAMPLE].str.rsplit(pat='_', n=1, expand=True)[0]
             chr_based_mutation_df[SAMPLE] = chr_based_mutation_df[SAMPLE].astype('category')
 
         if SAMPLE not in mutations_probabilities_df.columns.values:
@@ -3277,7 +3278,7 @@ def readChrBasedMutationsMergeWithProbabilitiesAndWrite(inputList):
 
         if simNum >= 1:
             # Get rid of simulation number at the end
-            chr_based_mutation_df[SAMPLE] = chr_based_mutation_df[SAMPLE].str.rsplit('_', 1, expand=True)[0]
+            chr_based_mutation_df[SAMPLE] = chr_based_mutation_df[SAMPLE].str.rsplit(pat='_', n=1, expand=True)[0]
 
         if (sigprofiler_simulator_mutation_context in SBS_CONTEXTS):
             chrBasedMergedMutationsFileName = 'chr%s_%s_for_topography.txt' %(chrShort, SUBS)

SigProfilerTopography/source/commons/TopographyCommons.py

@@ -34,6 +34,9 @@
 
 from matplotlib import pyplot as plt
 
+import matplotlib.cm as matplotlib_cm
+from matplotlib import colors
+
 import matplotlib as mpl
 import matplotlib.cm as cm
 from matplotlib.colors import Normalize
@@ -59,6 +62,12 @@
 MINIMUM_REQUIRED_PROCESSIVE_GROUP_LENGTH = 2
 MINIMUM_REQUIRED_NUMBER_OF_PROCESSIVE_GROUPS = 2
 
+def truncate_colormap(cmap, minval=0.0, maxval=1.0, n=100):
+    new_cmap = colors.LinearSegmentedColormap.from_list(
+        'trunc({n},{a:.2f},{b:.2f})'.format(n=cmap.name, a=minval, b=maxval),
+        cmap(np.linspace(minval, maxval, n)))
+    return new_cmap
+
 def readSimulationBasedDictionaries(outputDir,jobname,numberofSimulations):
     simulation2Signature2ProcessiveGroupLength2PropertiesDict = {}
 
@@ -457,11 +466,12 @@ def plot_processivity_figure(outputDir,
 
         panel1.set_yticks(np.arange(0, len(sorted_signature_list) + 1, 1))
 
-        cmap = cm.get_cmap('YlOrRd')  # Looks better good
+        cmap = matplotlib_cm.get_cmap('Purples')
+        cmap = truncate_colormap(cmap, 0.2, 1)
+
         v_min = 2
         v_max = 20
-        # Very important: You have to normalize
-        norm = mpl.colors.Normalize(vmin=v_min, vmax=v_max)
+        norm = plt.Normalize(v_min, v_max) # legacy
 
         if not signature_processive_group_length_properties_df.empty:
             # Plot the circles with color

SigProfilerTopography/source/plotting/ProcessivityFigures.py

@@ -1,6 +1,6 @@
 
 # THIS FILE IS GENERATED FROM SIGPROFILERTOPOGRAPHY SETUP.PY
-short_version = '1.0.73'
-version = '1.0.73'
+short_version = '1.0.74'
+version = '1.0.74'
 
 

SigProfilerTopography/version.py

@@ -10,7 +10,7 @@ def readme():
 	with open('README.rst') as f:
 		return(f.read())
 
-VERSION = '1.0.73'
+VERSION = '1.0.74'
 
 def write_version_py(filename='SigProfilerTopography/version.py'):
 	# Copied from numpy setup.py