move legend to up and fix y-axis inward if using different ymax

deeptools/heatmapper.py

@@ -1308,9 +1308,9 @@ def hmcluster(self, k, evaluate_silhouette=True, method='kmeans', clustering_sam
         _clustered_matrix = []
         cluster_number = 1
         for cluster in cluster_order:
-            self.group_labels.append("cluster_{}".format(cluster_number))
-            cluster_number += 1
             cluster_ids = _cluster_ids_list[cluster]
+            self.group_labels.append("cluster_{}_n{}".format(cluster_number,len(cluster_ids)))
+            cluster_number += 1
             self.group_boundaries.append(self.group_boundaries[-1] +
                                          len(cluster_ids))
             _clustered_matrix.append(self.matrix[cluster_ids, :])

deeptools/heatmapper_utilities.py

@@ -9,6 +9,9 @@
 
 old_settings = np.seterr(all='ignore')
 
+debug = 0
+if debug:
+    from ipdb import set_trace
 
 def plot_single(ax, ma, average_type, color, label, plot_type='lines'):
     """
@@ -18,7 +21,7 @@ def plot_single(ax, ma, average_type, color, label, plot_type='lines'):
     ----------
     ax : matplotlib axis
         matplotlib axis
-    ma : numpy array
+    ma : numpy array or list of numpy array(for plot with --repgrplist, take average between replicates )
         numpy array The data on this matrix is summarized according
         to the `average_type` argument.
     average_type : str
@@ -32,7 +35,9 @@ def plot_single(ax, ma, average_type, color, label, plot_type='lines'):
         type of plot. Either 'se' for standard error, 'std' for
         standard deviation, 'overlapped_lines' to plot each line of the matrix,
         fill to plot the area between the x axis and the value or any other string to
-        just plot the average line.
+        just plot the average line. When assign samples to replicates group such as
+        '--repgrplist WT WT KO KO' : 'std' would be the standard deviation between replicates groups.
+        'se' for standard error between replicates groups.
 
     Returns
     -------
@@ -63,7 +68,15 @@ def plot_single(ax, ma, average_type, color, label, plot_type='lines'):
 
 
     """
-    summary = np.ma.__getattribute__(average_type)(ma, axis=0)
+    if isinstance(ma,list):
+        summary_list = []
+        for ma_each in ma:
+            tmp = np.ma.__getattribute__(average_type)(ma_each, axis=0)
+            summary_list.append(tmp)
+        ma = np.array(summary_list)
+        summary = np.ma.__getattribute__("average")(ma, axis=0)
+    else:
+        summary = np.ma.__getattribute__(average_type)(ma, axis=0)
     # only plot the average profiles without error regions
     x = np.arange(len(summary))
     if isinstance(color, np.ndarray):
@@ -72,6 +85,8 @@ def plot_single(ax, ma, average_type, color, label, plot_type='lines'):
     if plot_type == 'fill':
         ax.fill_between(x, summary, facecolor=color, alpha=0.6, edgecolor='none')
 
+    if debug:
+        set_trace()
     if plot_type in ['se', 'std']:
         if plot_type == 'se':  # standard error
             std = np.std(ma, axis=0) / np.sqrt(ma.shape[0])

deeptools/parserCommon.py

@@ -849,6 +849,14 @@ def heatmapperOptionalArgs(mode=['heatmap', 'profile'][0]):
                           'group.',
                           action='store_true')
 
+    optional.add_argument('--repgrplist',
+                          default=None,
+                          nargs='+',
+                          help='Group profiles by genotype'
+                               'assign each profile to a group(as replicates) to plot average +/- se/std.'
+                               'If the number of group values is smaller than'
+                               'the number of samples, the values will be equally divide into groups.')
+
     optional.add_argument('--plotFileFormat',
                           metavar='',
                           help='Image format type. If given, this '

deeptools/plotHeatmap.py

@@ -13,8 +13,8 @@
 from matplotlib.font_manager import FontProperties
 import matplotlib.gridspec as gridspec
 from matplotlib import ticker
-import copy
-import sys
+
+import sys, re, os, copy
 import plotly.offline as py
 import plotly.graph_objs as go
 
@@ -117,6 +117,21 @@ def prepare_layout(hm_matrix, heatmapsize, showSummaryPlot, showColorbar, perGro
 
     return grids
 
+def autobreaklinetitle(title,sep="[-_,.:]",lmax=15):
+    outsep = "-"
+    sss = [ rr for rr in re.split(sep,title) if len(rr) ]
+    newtitle, tmp = "", ""
+    for ss in sss:
+        tmp0 = tmp
+        tmp += ss
+        if len(tmp) > lmax:
+            newtitle += tmp0.strip(outsep) + "\n"
+            tmp = ss
+        else:
+            tmp += outsep
+    newtitle += tmp.strip(outsep)
+    newtitle = "\n" + newtitle
+    return newtitle
 
 def addProfilePlot(hm, plt, fig, grids, iterNum, iterNum2, perGroup, averageType, plot_type, yAxisLabel, color_list, yMin, yMax, wspace, hspace, colorbar_position, label_rotation=0.0):
     """
@@ -146,7 +161,7 @@ def addProfilePlot(hm, plt, fig, grids, iterNum, iterNum2, perGroup, averageType
             else:
                 ax_profile = fig.add_subplot(grids[0, sample_id])
 
-        ax_profile.set_title(title)
+        ax_profile.set_title(autobreaklinetitle(title))
         for group in range(iterNum2):
             if perGroup:
                 sub_matrix = hm.matrix.get_matrix(sample_id, group)
@@ -163,6 +178,7 @@ def addProfilePlot(hm, plt, fig, grids, iterNum, iterNum2, perGroup, averageType
         if sample_id > 0 and len(yMin) == 1 and len(yMax) == 1:
             plt.setp(ax_profile.get_yticklabels(), visible=False)
 
+        ax_profile.get_yaxis().set_tick_params(direction='in',pad=-22) # beisi
         if sample_id == 0 and yAxisLabel != '':
             ax_profile.set_ylabel(yAxisLabel)
         xticks, xtickslabel = hm.getTicks(tickIdx)
@@ -595,8 +611,13 @@ def plotMatrix(hm, outFileName,
             ax_list = addProfilePlot(hm, plt, fig, grids, iterNum, iterNum2, perGroup, averageType, plot_type, yAxisLabel, color_list, yMin, yMax, kwargs['wspace'], kwargs['hspace'], colorbar_position, label_rotation)
 
         if legend_location != 'none':
-            ax_list[-1].legend(loc=legend_location.replace('-', ' '), ncol=1, prop=fontP,
-                               frameon=False, markerscale=0.5)
+            ax = ax_list[-1] # beisi
+            box = ax.get_position()
+            ax.set_position([box.x0, box.y0 - box.height * 0.1, box.width, box.height * 0.9])
+            legend = ax.legend(loc='lower right', shadow=False, fontsize='x-large', bbox_to_anchor=(0, 1.3, 1, .22), ncol=10, frameon=False, prop=fontP) # beisi, legend line
+            ax.add_artist(legend)
+            #            ax_list[-1].legend(loc=legend_location.replace('-', ' '), ncol=1, prop=fontP,
+            #                   frameon=False, markerscale=0.5)
 
     first_group = 0  # helper variable to place the title per sample/group
     for sample in range(hm.matrix.get_num_samples()):
@@ -632,7 +653,7 @@ def plotMatrix(hm, outFileName,
 
             if group == first_group and not showSummaryPlot and not perGroup:
                 title = hm.matrix.sample_labels[sample]
-                ax.set_title(title)
+                ax.set_title(autobreaklinetitle(title))
 
             if box_around_heatmaps is False:
                 # Turn off the boxes around the individual heatmaps
@@ -685,9 +706,9 @@ def plotMatrix(hm, outFileName,
                 ax.axes.set_xlabel(xAxisLabel)
             ax.axes.set_yticks([])
             if perGroup and group == 0:
-                ax.axes.set_ylabel(sub_matrix['sample'])
+                ax.axes.set_ylabel(sub_matrix['sample'],rotation=75,labelpad=0,fontsize=15)
             elif not perGroup and sample == 0:
-                ax.axes.set_ylabel(sub_matrix['group'])
+                ax.axes.set_ylabel(sub_matrix['group'],rotation=75,labelpad=0,horizontalalignment='right',fontsize=15)
 
             # Plot vertical lines at tick marks if desired
             if linesAtTickMarks: