markernames labeling issues fixed in manhatten plot

README.md

@@ -374,7 +374,7 @@ ANOVA summary, multiple pairwise comparisons, and assumption tests statistics
 <b>Manhatten plot</b>
 
 `bioinfokit.visuz.marker.mhat(df, chr, pv, color, dim, r, ar, gwas_sign_line, gwasp, dotsize, markeridcol, markernames, 
-    gfont, valpha, show, figtype, axxlabel, axylabel, axlabelfontsize, ylm)`
+    gfont, valpha, show, figtype, axxlabel, axylabel, axlabelfontsize, ylm, gstyle)`
 
 Parameters | Description
 ------------ | -------------
@@ -386,8 +386,8 @@ Parameters | Description
 `gwasp` |  Statistical significant threshold to identify significant SNPs [float][default: 5E-08]
 `dotsize`| The size of the dots in the plot [float][default: 8]
 `markeridcol` | Name of a column having SNPs. This is necessary for plotting SNP names on the plot [string][default: None]
-`markernames` | The list of the SNPs to display on the plot. These SNP should be present in SNP column. Additionally, it also accepts the dict of SNPs and its associated gene name. If this option set to True, it will label all SNPs with P-value significant score defined by `gwasp` [string, list, dict][default: True]
-`gfont` | Font size for SNP names to display on the plot [float][default: 8]
+`markernames` | The list of the SNPs to display on the plot. These SNP should be present in SNP column. Additionally, it also accepts the dict of SNPs and its associated gene name. If this option set to True, it will label all SNPs with P-value significant score defined by `gwasp` [string, list, tuple, dict][default: True]
+`gfont` | Font size for SNP names to display on the plot [float][default: 8].  gfont not compatible with gstyle=2.
 `valpha` | Transparency of points on plot [float (between 0 and 1)][default: 1.0]
 `dim` | Figure size [tuple of two floats (width, height) in inches][default: (6, 4)]
 `r` | Figure resolution in dpi [int][default: 300]
@@ -398,6 +398,8 @@ Parameters | Description
 `axylabel` | Label for Y-axis. If you provide this option, default label will be replaced [string][default: None]
 `axlabelfontsize` | Font size for axis labels [float][default: 9]
 `ylm` | Range of ticks to plot on Y-axis [float tuple (bottom, top, interval)][default: None]
+`gstyle` | Style of the text for markernames. 1 for default text and 2 for box text [int][default: 1]
+
 
 
 Returns:

VERSIONLOG.md

@@ -1,3 +1,8 @@
+v0.7.3 has the following updates and changes (May 14, 2020)
+- In manhatten plot (`visuz.marker.mhat`), the labeling issue with `markernames` parameter corrected (see issue # 4 on github for details;
+  thanks to mkirchler for reporting)
+- `gstyle` parameter added in manhatten plot for box style annotation
+
 v0.7.2 has the following updates and changes (May 08, 2020)
 - `splitvcf` function added for splitting VCF file into individual VCF files for each chromosome
 - `mergevcf` moved to `analys.marker` class

bioinfokit/__init__.py

@@ -1,5 +1,5 @@
 name = "bioinfokit"
-__version__ = "0.7.2"
+__version__ = "0.7.3"
 __author__ = "Renesh Bedre"
 
 

bioinfokit/visuz.py

@@ -77,7 +77,9 @@ def venn(vennset=(1,1,1,1,1,1,1), venncolor=('#00909e', '#f67280', '#ff971d'), v
     else:
         print("Error: check the set dataset")
 
-class gene_exp():
+
+class gene_exp:
+
     def __init__(self):
         pass
 
@@ -267,31 +269,55 @@ def depr_mes(func_name):
         print("This function is deprecated. Please use", func_name )
         print("Read docs at https://reneshbedre.github.io/blog/howtoinstall.html")
 
-class marker():
-    def geneplot_mhat(df, markeridcol, chr, pv, gwasp, markernames, gfont, ax):
+
+class marker:
+
+    def __init__(self):
+        pass
+
+    def geneplot_mhat(df, markeridcol, chr, pv, gwasp, markernames, gfont, gstyle, ax):
         if markeridcol is not None:
             if markernames is not None and markernames is True:
                 for i in df[markeridcol].unique():
                     if df.loc[df[markeridcol] == i, pv].iloc[0] <= gwasp:
-                        plt.text((df.loc[df[markeridcol] == i, 'ind'].iloc[0]), df.loc[df[markeridcol] == i, 'tpval'].iloc[0],
-                                 str(i), fontsize=gfont)
-            elif markernames is not None and type(markernames) is tuple:
+                        if gstyle == 1:
+                            plt.text(df.loc[df[markeridcol] == i, 'ind'].iloc[0], df.loc[df[markeridcol] == i, 'tpval'].iloc[0],
+                                    str(i), fontsize=gfont)
+                        elif gstyle == 2:
+                            plt.annotate(i, xy=(df.loc[df[markeridcol] == i, 'ind'].iloc[0], df.loc[df[markeridcol] == i, 'tpval'].iloc[0]),
+                                         xycoords='data', xytext=(5, -15), textcoords='offset points', size=6,
+                                         bbox=dict(boxstyle="round", alpha=0.2),
+                                         arrowprops=dict(arrowstyle="wedge,tail_width=0.5", alpha=0.2, relpos=(0, 0)))
+            elif markernames is not None and isinstance(markernames, (tuple, list)):
                 for i in df[markeridcol].unique():
-                    plt.text(df.loc[df[markeridcol] == i, 'ind'].iloc[0], df.loc[df[markeridcol] == i, 'tpval'].iloc[0],
-                                 str(i), fontsize=gfont)
-            elif markernames is not None and type(markernames) is dict:
+                    if i in markernames:
+                        if gstyle == 1:
+                            plt.text(df.loc[df[markeridcol] == i, 'ind'].iloc[0], df.loc[df[markeridcol] == i, 'tpval'].iloc[0],
+                                str(i), fontsize=gfont)
+                        elif gstyle == 2:
+                            plt.annotate(i, xy=(df.loc[df[markeridcol] == i, 'ind'].iloc[0], df.loc[df[markeridcol] == i, 'tpval'].iloc[0]),
+                                         xycoords='data', xytext=(5, -15), textcoords='offset points', size=6,
+                                         bbox=dict(boxstyle="round", alpha=0.2),
+                                         arrowprops=dict(arrowstyle="wedge,tail_width=0.5", alpha=0.2, relpos=(0, 0)))
+            elif markernames is not None and isinstance(markernames, dict):
                 for i in df[markeridcol].unique():
                     if i in markernames:
-                       plt.text(df.loc[df[markeridcol] == i, 'ind'].iloc[0], df.loc[df[markeridcol] == i, 'tpval'].iloc[0],
+                        if gstyle == 1:
+                            plt.text(df.loc[df[markeridcol] == i, 'ind'].iloc[0], df.loc[df[markeridcol] == i, 'tpval'].iloc[0],
                                  markernames[i], fontsize=gfont)
+                        elif gstyle == 2:
+                            plt.annotate(markernames[i], xy=(
+                            df.loc[df[markeridcol] == i, 'ind'].iloc[0], df.loc[df[markeridcol] == i, 'tpval'].iloc[0]),
+                                         xycoords='data', xytext=(5, -15), textcoords='offset points', size=6,
+                                         bbox=dict(boxstyle="round", alpha=0.2),
+                                         arrowprops=dict(arrowstyle="wedge,tail_width=0.5", alpha=0.2, relpos=(0, 0)))
         else:
-            print("Error: provide 'markeridcol' parameter")
-            sys.exit(1)
+            raise Exception("provide 'markeridcol' parameter")
 
     def mhat(df="dataframe", chr=None, pv=None, color=None, dim=(6,4), r=300, ar=90, gwas_sign_line=False,
              gwasp=5E-08, dotsize=8, markeridcol=None, markernames=None, gfont=8, valpha=1, show=False, figtype='png',
              axxlabel=None, axylabel=None, axlabelfontsize=9, axlabelfontname="Arial", axtickfontsize=9,
-             axtickfontname="Arial", ylm=None):
+             axtickfontname="Arial", ylm=None, gstyle=1):
 
         _x, _y = 'Chromosomes', r'$ -log_{10}(P)$'
         rand_colors = ('#a7414a', '#282726', '#6a8a82', '#a37c27', '#563838', '#0584f2', '#f28a30', '#f05837',
@@ -316,7 +342,7 @@ def mhat(df="dataframe", chr=None, pv=None, color=None, dim=(6,4), r=300, ar=90,
             if df[chr].nunique() % 2 == 0:
                 color_list = list(reduce(lambda x, y: x+y, zip(color_1, color_2)))
             elif df[chr].nunique() % 2 == 1:
-                color_list = list(reduce(lambda x, y: x + y, zip(color_1, color_2)))
+                color_list = list(reduce(lambda x, y: x+y, zip(color_1, color_2)))
                 color_list.append(color[0])
         elif color is not None and len(color) == df[chr].nunique():
             color_list = color
@@ -343,7 +369,7 @@ def mhat(df="dataframe", chr=None, pv=None, color=None, dim=(6,4), r=300, ar=90,
         if gwas_sign_line is True:
             ax.axhline(y=-np.log10(gwasp), linestyle='--', color='#7d7d7d', linewidth=1)
         if markernames is not None:
-            marker.geneplot_mhat(df, markeridcol, chr, pv, gwasp, markernames, gfont, ax=ax)
+            marker.geneplot_mhat(df, markeridcol, chr, pv, gwasp, markernames, gfont, gstyle, ax=ax)
         ax.margins(x=0)
         ax.margins(y=0)
         ax.set_xticks(xticks)