gwaslab 3.4.48__py3-none-any.whl → 3.5.0__py3-none-any.whl

gwaslab/viz_plot_compare_effect.py

@@ -9,6 +9,7 @@ from matplotlib.patches import Rectangle
 from adjustText import adjust_text
 from gwaslab.viz_aux_save_figure import save_figure
 from gwaslab.util_in_get_sig import getsig
+from gwaslab.util_in_get_sig import annogene
 from gwaslab.g_Log import Log
 from gwaslab.util_in_correct_winnerscurse import wc_correct
 from gwaslab.util_in_correct_winnerscurse import wc_correct_test
@@ -59,6 +60,7 @@ def compare_effect(path1,
                    xylabel_prefix="Per-allele effect size in ",
                    helper_line_args=None,
                    fontargs=None,
+                   build="19",
                    r_or_r2="r",
                    # 
                    errargs=None,
@@ -77,10 +79,9 @@ def compare_effect(path1,
         scaled2 = True
     if is_q_mc=="fdr" or is_q_mc=="bon":
         is_q = True
-    
     if is_q == True:
         if is_q_mc not in [False,"fdr","bon","non"]:
-            raise ValueError("Please select either fdr or bon or non for is_q_mc.")
+            raise ValueError('Please select either "fdr" or "bon" or "non"/False for is_q_mc.')
     if save_args is None:
         save_args = {"dpi":300,"facecolor":"white"}
     if reg_box is None:
@@ -89,6 +90,8 @@ def compare_effect(path1,
         sep = ["\t","\t"]
     if get_lead_args is None:
         get_lead_args = {}
+    if anno=="GENENAME":
+        get_lead_args["anno"]=True
     if errargs is None:
         errargs={"ecolor":"#cccccc","elinewidth":1}
     if fontargs is None:
@@ -191,10 +194,12 @@ def compare_effect(path1,
         ######### 8.1 if a snplist is provided, use the snp list
         log.write(" -Extract variants in the given list from "+label[0]+"...")
         sig_list_1 = sumstats.loc[sumstats["SNPID"].isin(snplist),:].copy()
+        if anno=="GENENAME":
+            sig_list_1 = annogene(sumstats,"SNPID","CHR","POS", build=build, verbose=verbose,**get_lead_args)
     else:
-        ######### 8,2 otherwise use the sutomatically detected lead SNPs
+        ######### 8,2 otherwise use the automatically detected lead SNPs
         log.write(" -Extract lead variants from "+label[0]+"...")
-        sig_list_1 = getsig(sumstats,"SNPID","CHR","POS","P", verbose=verbose,sig_level=sig_level,**get_lead_args)
+        sig_list_1 = getsig(sumstats,"SNPID","CHR","POS","P", build=build, verbose=verbose,sig_level=sig_level,**get_lead_args)
     
     if drop==True:
         sig_list_1 = drop_duplicate_and_na(sig_list_1, sort_by="P", log=log ,verbose=verbose)
@@ -235,10 +240,12 @@ def compare_effect(path1,
         ######### 12.1 if a snplist is provided, use the snp list
         log.write(" -Extract snps in the given list from "+label[1]+"...")
         sig_list_2 = sumstats.loc[sumstats["SNPID"].isin(snplist),:].copy()
+        if anno=="GENENAME":
+            sig_list_2 = annogene(sumstats,"SNPID","CHR","POS", build=build, verbose=verbose,**get_lead_args)
     else: 
         log.write(" -Extract lead snps from "+label[1]+"...")
         ######### 12.2 otherwise use the sutomatically detected lead SNPs
-        sig_list_2 = getsig(sumstats,"SNPID","CHR","POS","P",
+        sig_list_2 = getsig(sumstats,"SNPID","CHR","POS","P",build=build,
                                  verbose=verbose,sig_level=sig_level,**get_lead_args)
     if drop==True:
         sig_list_2 = drop_duplicate_and_na(sig_list_2, sort_by="P", log=log ,verbose=verbose)
@@ -248,6 +255,10 @@ def compare_effect(path1,
     log.write("Merging snps from "+label[0]+" and "+label[1]+"...")
     
     sig_list_merged = pd.merge(sig_list_1,sig_list_2,left_on="SNPID",right_on="SNPID",how="outer",suffixes=('_1', '_2'))
+    if anno == "GENENAME":
+        sig_list_merged.loc[sig_list_merged["SNPID"].isin((sig_list_1["SNPID"])),"GENENAME"] = sig_list_merged.loc[sig_list_merged["SNPID"].isin((sig_list_1["SNPID"])),"GENE_1"]
+        sig_list_merged.loc[~sig_list_merged["SNPID"].isin((sig_list_1["SNPID"])),"GENENAME"] = sig_list_merged.loc[~sig_list_merged["SNPID"].isin((sig_list_1["SNPID"])),"GENE_2"]
+        sig_list_merged = sig_list_merged.drop(columns=["GENE_1","GENE_2","LOCATION_1","LOCATION_2"])
     #     SNPID       P_1       P_2
     #0   rs117986209  0.142569  0.394455
     #1     rs6704312  0.652104  0.143750
@@ -533,7 +544,7 @@ def compare_effect(path1,
 
     ########################## Het test############################################################
     ## heterogeneity test
-    if (is_q is True):
+    if (is_q == True):
         log.write(" -Calculating Cochran's Q statistics and peform chisq test...", verbose=verbose)
         if mode=="beta" or mode=="BETA" or mode=="Beta":
             sig_list_merged = test_q(sig_list_merged,"EFFECT_1","SE_1","EFFECT_2_aligned","SE_2",q_level=q_level,is_q_mc=is_q_mc, log=log, verbose=verbose)
@@ -552,7 +563,7 @@ def compare_effect(path1,
         log.write(" -Exclude "+str(len(sig_list_merged) -sum(both_eaf_clear))+ " variants with maf <",maf_level, verbose=verbose)
         sig_list_merged = sig_list_merged.loc[both_eaf_clear,:]
     # heterogeneity summary
-    if (is_q is True):
+    if (is_q == True):
         log.write(" -Significant het:" ,len(sig_list_merged.loc[sig_list_merged["HetP"]<0.05,:]), verbose=verbose)
         log.write(" -All sig:" ,len(sig_list_merged), verbose=verbose)
         log.write(" -Het rate:" ,len(sig_list_merged.loc[sig_list_merged["HetP"]<0.05,:])/len(sig_list_merged), verbose=verbose)   
@@ -633,11 +644,11 @@ def compare_effect(path1,
             ax.scatter(both["OR_1"],both["OR_2_aligned"],label=label[2],zorder=2,color="#205be6",edgecolors=both["Edge_color"],marker="s",**scatterargs)
             legend_elements.append(label[2])
     ## annotation #################################################################################################################
-    if anno==True:
+    if anno==True or anno=="GENENAME":
         sig_list_toanno = sig_list_merged.dropna(axis=0)
         if is_q==True and anno_het == True:
             sig_list_toanno = sig_list_toanno.loc[sig_list_toanno["Edge_color"]=="black",:]
-        
+
         if mode=="beta":
             sig_list_toanno = sig_list_toanno.loc[sig_list_toanno["EFFECT_1"].abs() >=anno_min1 ,:]
             sig_list_toanno = sig_list_toanno.loc[sig_list_toanno["EFFECT_2_aligned"].abs() >=anno_min2 ,:]
@@ -651,22 +662,38 @@ def compare_effect(path1,
 
         texts_l=[]
         texts_r=[]
+        
+        if anno==True:
+            log.write("Annotating variants using {}".format("SNPID"), verbose=verbose)
+        elif anno=="GENENAME":
+            log.write("Annotating variants using {}".format("GENENAME"), verbose=verbose)
+        
         for index, row in sig_list_toanno.iterrows():
+            log.write("Annotating {}...".format(row), verbose=verbose)
+            if anno==True:
+                to_anno_text = index
+            elif type(anno) is str:
+                if not pd.isna(row[anno]):
+                    to_anno_text = row[anno]
+                else:
+                    to_anno_text = index
+
             if mode=="beta" or mode=="BETA" or mode=="Beta":
                 if row["EFFECT_1"] <  row["EFFECT_2_aligned"]:
-                    texts_l.append(plt.text(row["EFFECT_1"], row["EFFECT_2_aligned"],index,ha="right",va="bottom"))
+                    texts_l.append(plt.text(row["EFFECT_1"], row["EFFECT_2_aligned"],to_anno_text,ha="right",va="bottom"))
                 else:
-                    texts_r.append(plt.text(row["EFFECT_1"], row["EFFECT_2_aligned"],index,ha="left",va="top"))
+                    texts_r.append(plt.text(row["EFFECT_1"], row["EFFECT_2_aligned"],to_anno_text,ha="left",va="top"))
             else:
                 if row["OR_1"] <  row["OR_2_aligned"]:
-                    texts_l.append(plt.text(row["OR_1"], row["OR_2_aligned"],index, ha='right', va='bottom')) 
+                    texts_l.append(plt.text(row["OR_1"], row["OR_2_aligned"],to_anno_text, ha='right', va='bottom')) 
                 else:
-                    texts_r.append(plt.text(row["OR_1"], row["OR_2_aligned"],index, ha='left', va='top')) 
-
-        adjust_text(texts_l,autoalign =False,precision =0.001,lim=1000, ha="right",va="bottom", expand_text=(1,1.8) , expand_objects=(0.1,0.1), expand_points=(1.8,1.8) ,force_objects=(0.8,0.8) ,arrowprops=dict(arrowstyle='-|>', color='grey'),ax=ax)
-        adjust_text(texts_r,autoalign =False,precision =0.001,lim=1000, ha="left",va="top", expand_text=(1,1.8) , expand_objects=(0.1,0.1), expand_points=(1.8,1.8) ,force_objects =(0.8,0.8),arrowprops=dict(arrowstyle='-|>', color='grey'),ax=ax)
-    
+                    texts_r.append(plt.text(row["OR_1"], row["OR_2_aligned"],to_anno_text, ha='left', va='top')) 
+        if len(texts_l)>0:
+            adjust_text(texts_l,autoalign =False,precision =0.001,lim=1000, ha="right",va="bottom", expand_text=(1,1.8) , expand_objects=(0.1,0.1), expand_points=(1.8,1.8) ,force_objects=(0.8,0.8) ,arrowprops=dict(arrowstyle='-|>', color='grey'),ax=ax)
+        if len(texts_r)>0:
+            adjust_text(texts_r,autoalign =False,precision =0.001,lim=1000, ha="left",va="top", expand_text=(1,1.8) , expand_objects=(0.1,0.1), expand_points=(1.8,1.8) ,force_objects =(0.8,0.8),arrowprops=dict(arrowstyle='-|>', color='grey'),ax=ax)
     elif type(anno) is dict:
+        sig_list_toanno = sig_list_merged.dropna(axis=0)
         # if input is a dict
         sig_list_toanno = sig_list_toanno.loc[sig_list_toanno.index.isin(list(anno.keys())),:]
         if is_q==True and anno_het == True:
@@ -696,9 +723,10 @@ def compare_effect(path1,
                     texts_l.append(plt.text(row["OR_1"], row["OR_2_aligned"],anno[index], ha='right', va='bottom')) 
                 else:
                     texts_r.append(plt.text(row["OR_1"], row["OR_2_aligned"],anno[index], ha='left', va='top')) 
-
-        adjust_text(texts_l,autoalign =False,precision =0.001,lim=1000, ha="right",va="bottom", expand_text=(1,1.8) , expand_objects=(0.1,0.1), expand_points=(1.8,1.8) ,force_objects=(0.8,0.8) ,arrowprops=dict(arrowstyle='-|>', color='grey'),ax=ax)
-        adjust_text(texts_r,autoalign =False,precision =0.001,lim=1000, ha="left",va="top", expand_text=(1,1.8) , expand_objects=(0.1,0.1), expand_points=(1.8,1.8) ,force_objects =(0.8,0.8),arrowprops=dict(arrowstyle='-|>', color='grey'),ax=ax)
+        if len(texts_l)>0:
+            adjust_text(texts_l,autoalign =False,precision =0.001,lim=1000, ha="right",va="bottom", expand_text=(1,1.8) , expand_objects=(0.1,0.1), expand_points=(1.8,1.8) ,force_objects=(0.8,0.8) ,arrowprops=dict(arrowstyle='-|>', color='grey'),ax=ax)
+        if len(texts_r)>0:
+            adjust_text(texts_r,autoalign =False,precision =0.001,lim=1000, ha="left",va="top", expand_text=(1,1.8) , expand_objects=(0.1,0.1), expand_points=(1.8,1.8) ,force_objects =(0.8,0.8),arrowprops=dict(arrowstyle='-|>', color='grey'),ax=ax)
     #################################################################################################################################
     
     # plot x=0,y=0, and a 45 degree line

gwaslab/viz_plot_miamiplot2.py

@@ -290,7 +290,11 @@ def plot_miami2(
     
 
     #####################################################################################################################    
-    
+    ax1l, ax1r = ax5.get_xlim()
+    ax5l, ax5r = ax1.get_xlim()
+    ax1.set_xlim([min(ax1l,ax5l), max(ax1r,ax5r)])
+    ax5.set_xlim([min(ax1l,ax5l), max(ax1r,ax5r)])
+    #####################################################################################################################
     ax5.set_xlabel("")
     #ax5.set_xticks(chrom_df)
     ax5.set_xticklabels([])

gwaslab/viz_plot_mqqplot.py

@@ -139,7 +139,7 @@ def mqqplot(insumstats,
           anno_gtf_path=None,
           anno_adjust=False,
           anno_max_iter=100,
-          arm_offset=50,
+          arm_offset=None,
           arm_scale=1,
           anno_height=1,
           arm_scale_d=None,
@@ -291,7 +291,7 @@ def mqqplot(insumstats,
     if maf_bin_colors is None:
         maf_bin_colors = ["#f0ad4e","#5cb85c", "#5bc0de","#000042"]
     if save_args is None:
-        save_args = {"dpi":300,"facecolor":"white"}
+        save_args = {"dpi":400,"facecolor":"white"}
     if highlight is None:
         highlight = list()
     if highlight_anno_args is None:
@@ -329,6 +329,20 @@ def mqqplot(insumstats,
                     fig_args["dpi"]=72
                     scatter_args["rasterized"]=True
                     qq_scatter_args["rasterized"]=True
+                else:
+                    fig_args["dpi"] = save_args["dpi"]
+    
+    # configure dpi if saving the plot
+    fig_args, scatter_args, qq_scatter_args, save_args = _configure_fig_save_kwargs(save = save, 
+                                                                                    fig_args = fig_args, 
+                                                                                    scatter_args = scatter_args, 
+                                                                                    qq_scatter_args = qq_scatter_args, 
+                                                                                    save_args = save_args)
+
+
+    if len(anno_d) > 0 and arm_offset is None:
+        # in pixels
+        arm_offset = fig_args["dpi"] * repel_force * fig_args["figsize"][0]*0.5
 
     log.write("Start to create MQQ plot...{}:".format(_get_version()),verbose=verbose)
     log.write(" -Genomic coordinates version: {}...".format(build),verbose=verbose)
@@ -401,7 +415,7 @@ def mqqplot(insumstats,
     if mode=="b":
         sig_level=1,
         sig_line=False,
-        windowsizekb = 100000000   
+        #windowsizekb = 100000000   
         mode="mb"
         scatter_args={"marker":"s"}
         marker_size= (marker_size[1],marker_size[1])
@@ -522,8 +536,12 @@ def mqqplot(insumstats,
                                                     pos=pos, 
                                                     verbose=verbose, 
                                                     log=log)
+        
+        lines_to_plot = pd.Series(lines_to_plot.to_list() + [bmean, bmedian])
+    
     else:
         bmean, bmedian=0,0 
+    
 # P value conversion #####################################################################################################  
     
     # add raw_P and scaled_P
@@ -956,7 +974,7 @@ def mqqplot(insumstats,
             ax1.set_title(mtitle,fontsize=title_fontsize,family=font_family)
         log.write("Finished processing figure arts.",verbose=verbose)
 
-        # Add annotation arrows and texts
+        ## Add annotation arrows and texts
         log.write("Start to annotate variants...",verbose=verbose)
         ax1 = annotate_single(
                                 sumstats=sumstats,
@@ -1055,7 +1073,8 @@ def mqqplot(insumstats,
         fig.suptitle(title , fontsize = title_fontsize ,x=0.5, y=1.05)
     else:
         fig.suptitle(title , fontsize = title_fontsize, x=0.5,y=1)
-
+        ## Add annotation arrows and texts
+    
     # Saving figure
     save_figure(fig = fig, save = save, keyword=mode, save_args=save_args, log = log, verbose=verbose)
 
@@ -1069,7 +1088,31 @@ def mqqplot(insumstats,
 
 ##############################################################################################################################################################################
 
+def _configure_fig_save_kwargs(save=None, 
+                               fig_args=None, 
+                               scatter_args=None, 
+                               qq_scatter_args=None, 
+                               save_args=None):
+    if fig_args is None:
+        fig_args = dict()
+    if scatter_args is None:
+        scatter_args = dict()
+    if qq_scatter_args is None:
+        qq_scatter_args = dict()
+    if save_args is None:
+        save_args = dict()
 
+    if save is not None:
+        if type(save) is not bool:
+            if len(save)>3:
+                if save[-3:]=="pdf" or save[-3:]=="svg":
+                    # to save as vectorized plot
+                    fig_args["dpi"]=72
+                    scatter_args["rasterized"]=True
+                    qq_scatter_args["rasterized"]=True
+                else:
+                    fig_args["dpi"] = save_args["dpi"]
+    return fig_args, scatter_args, qq_scatter_args, save_args
 
 
 def _add_pad_to_x_axis(ax1, xpad, xpadl, xpadr, sumstats, pos, chrpad, xtight, log, verbose):
@@ -1104,12 +1147,6 @@ def _add_pad_to_x_axis(ax1, xpad, xpadl, xpadr, sumstats, pos, chrpad, xtight, l
     return ax1
 
 
-
-
-
-
-
-
 ##############################################################################################################################################################################
 def _configure_cols_to_use(insumstats, snpid,  chrom, pos, ea, nea, eaf, p, mlog10p,scaled, mode,stratified,anno, anno_set, anno_alias,_chrom_df_for_i,highlight ,pinpoint,density_color):
     usecols=[]
@@ -1287,9 +1324,10 @@ def _process_density(sumstats, mode, bwindowsizekb, chrom, pos, verbose, log):
                 else:
                     break
         df = pd.DataFrame(stack,columns=["SNPID","TCHR+POS","DENSITY"])
-        sumstats["DENSITY"] = df["DENSITY"].values
-        bmean=sumstats["DENSITY"].mean()
-        bmedian=sumstats["DENSITY"].median()
+        sumstats["DENSITY"] = df["DENSITY"].astype("Float64").values
+
+        bmean=sumstats.drop_duplicates(subset="SNPID")["DENSITY"].mean()
+        bmedian=sumstats.drop_duplicates(subset="SNPID")["DENSITY"].median()
     elif "b" in mode and "DENSITY" in sumstats.columns:
         bmean=sumstats["DENSITY"].mean()
         bmedian=sumstats["DENSITY"].median()
@@ -1305,6 +1343,7 @@ def _process_line(ax1, sig_line, suggestive_sig_line, additional_line, lines_to_
                                 linestyle="--",
                                 color=sig_line_color,
                                 zorder=1)
+        
     if suggestive_sig_line is True:
         suggestive_sig_line = ax1.axhline(y=lines_to_plot[1], 
                                             linewidth = sc_linewidth, 
@@ -1312,15 +1351,20 @@ def _process_line(ax1, sig_line, suggestive_sig_line, additional_line, lines_to_
                                             color=suggestive_sig_line_color,
                                             zorder=1)
     if additional_line is not None:
-        for index, level in enumerate(lines_to_plot[2:].values):
+        for index, level in enumerate(lines_to_plot[2:2+len(additional_line)].values):
             ax1.axhline(y=level, 
                         linewidth = sc_linewidth, 
                         linestyle="--", 
                         color=additional_line_color[index%len(additional_line_color)],
                         zorder=1)
-    if "b" in mode:   
+    if "b" in mode: 
+        bmean = lines_to_plot.iat[-2]
+        bmedian = lines_to_plot.iat[-1]
         # for brisbane plot, add median and mean line 
+        log.write(" -Plotting horizontal line (  mean DENISTY): y = {}".format(bmean),verbose=verbose) 
         meanline = ax1.axhline(y=bmean, linewidth = sc_linewidth,linestyle="-",color=sig_line_color,zorder=1000)
+
+        log.write(" -Plotting horizontal line ( median DENISTY): y = {}".format(bmedian),verbose=verbose) 
         medianline = ax1.axhline(y=bmedian, linewidth = sc_linewidth,linestyle="--",color=sig_line_color,zorder=1000)
     return ax1
 
@@ -1441,10 +1485,16 @@ def _process_layout(mode, figax, fig_args, mqqratio, region_hspace):
             ax2 = None
             plt.subplots_adjust(hspace=region_hspace)
     elif mode =="b" :
-        fig_args["figsize"] = (15,5)
-        fig, ax1 = plt.subplots(1, 1,**fig_args)
-        ax2 = None
-        ax3 = None
+        if figax is not None:
+            fig = figax[0]
+            ax1 = figax[1]
+            ax3 = None
+            ax2 = None
+        else:
+            fig_args["figsize"] = (15,5)
+            fig, ax1 = plt.subplots(1, 1,**fig_args)
+            ax2 = None
+            ax3 = None
     else:
         raise ValueError("Please select one from the 5 modes: mqq/qqm/m/qq/r/b")
     ax4=None

gwaslab/viz_plot_phe_heatmap.py

@@ -0,0 +1,260 @@
+import pandas as pd
+import matplotlib.pyplot as plt
+import seaborn as sns
+import numpy as np
+import scipy as sp
+from gwaslab.viz_aux_quickfix import _quick_assign_i_with_rank
+from gwaslab.viz_aux_quickfix import _get_largenumber
+from gwaslab.viz_aux_quickfix import _quick_fix_p_value
+from gwaslab.viz_aux_quickfix import _quick_fix_pos
+from gwaslab.viz_aux_quickfix import _quick_fix_chr
+from gwaslab.viz_aux_quickfix import _quick_fix_eaf
+from gwaslab.viz_aux_quickfix import _quick_fix_mlog10p
+from gwaslab.viz_aux_quickfix import _dropna_in_cols
+from gwaslab.viz_plot_mqqplot import _process_p_value
+from gwaslab.viz_plot_mqqplot import _configure_fig_save_kwargs
+from gwaslab.viz_plot_mqqplot import mqqplot
+from gwaslab.viz_aux_save_figure import save_figure
+from gwaslab.g_Log import Log
+import copy
+from gwaslab.bd_common_data import get_chr_to_number
+from gwaslab.bd_common_data import get_number_to_chr
+from gwaslab.g_version import _get_version
+
+def _gwheatmap(
+    insumstats,    
+    chrom="CHR",
+    pos="POS",
+    ref_chrom="REF_CHR",
+    ref_pos="REF_START",
+    p="P",
+    scaled=False,
+    sizes = (10,50),
+    alpha=0.5,
+    mlog10p="MLOG10P",
+    snpid="SNPID",
+    eaf=None,
+    group="CIS/TRANS",
+    ea="EA",
+    nea="NEA",
+    colors=None,
+    check = True,
+    chr_dict = None,
+    xchrpad = 0, 
+    ychrpad=0,
+    use_rank = False,
+    xtick_chr_dict=None, 
+    ytick_chr_dict=None, 
+    fontsize=10, 
+    add_b =False,
+    log=Log(),
+    fig_kwargs=None,
+    scatter_kwargs=None,
+    height_ratios=None,
+    hspace = 0.1,
+    font_family="Arial",
+    cis_windowsizekb=100,
+    verbose=True,
+    save=True,
+    save_kwargs=None,
+    grid_linewidth=1,
+    grid_linecolor="grey",
+    **mqq_kwargs
+):
+    log.write("Start to create genome-wide scatter plot...{}:".format(_get_version()),verbose=verbose)
+    if height_ratios is None:
+        height_ratios = [1, 2]
+    if xtick_chr_dict is None:         
+        xtick_chr_dict = get_number_to_chr()
+    if ytick_chr_dict is None:         
+        ytick_chr_dict = get_number_to_chr()
+    if chr_dict is None:          
+        chr_dict = get_chr_to_number()
+    if colors is None:
+        colors=["#CB132D","#597FBD"]
+    if fig_kwargs is None:
+        fig_kwargs= dict(figsize=(15,15))
+    if save_kwargs is None:
+        save_kwargs = {"dpi":300,"facecolor":"white"}
+    if scatter_kwargs is None:
+        scatter_kwargs = {}
+
+    fig_kwargs, scatter_kwargs, qq_scatter_args, save_kwargs = _configure_fig_save_kwargs(save=save, 
+                                                                                    fig_args = fig_kwargs, 
+                                                                                    scatter_args = scatter_kwargs, 
+                                                                                    qq_scatter_args = dict(), 
+                                                                                    save_args = save_kwargs)
+
+    sumstats = insumstats.copy()
+    
+    # Data QC and format
+    if check ==True:
+        sumstats[pos] = _quick_fix_pos(sumstats[pos])
+        sumstats[chrom] = _quick_fix_chr(sumstats[chrom], chr_dict=chr_dict)
+        sumstats[ref_pos] = _quick_fix_pos(sumstats[ref_pos])
+        sumstats[ref_chrom] = _quick_fix_chr(sumstats[ref_chrom], chr_dict=chr_dict)
+        sumstats = _dropna_in_cols(sumstats, [pos, chrom, ref_pos, ref_chrom], log=log, verbose=verbose)
+    
+    # dropna
+    sumstats = sumstats.sort_values(by=group)
+
+    if scaled is True:
+        sumstats["raw_P"] = pd.to_numeric(sumstats[mlog10p], errors='coerce')
+    else:
+        sumstats["raw_P"] = sumstats[p].astype("float64")
+
+    sumstats =  _process_p_value(sumstats=sumstats, 
+                                mode="m",
+                                p=p, 
+                                mlog10p=mlog10p, 
+                                scaled=scaled, 
+                                log=log, 
+                                verbose=verbose )
+    
+    
+    
+    if add_b ==False:
+        fig, ax1 = plt.subplots(**fig_kwargs)
+    else:
+        fig, (ax2, ax1) = plt.subplots( nrows=2 ,sharex=True, gridspec_kw={'height_ratios': height_ratios }, **fig_kwargs)
+        plt.subplots_adjust(hspace=hspace)   
+
+    ## assign i for variants
+    sumstats, chrom_df_x = _quick_assign_i_with_rank(sumstats, 
+                                                chrpad=xchrpad, 
+                                                use_rank=use_rank, 
+                                                chrom=chrom,
+                                                pos=pos,
+                                                verbose=verbose)
+    chrom_df_b = chrom_df_x
+    sumstats = sumstats.rename(columns={"i":"i_x"})
+    add_x_unique = list(sumstats["_ADD"].unique()) 
+    
+    ## determine grouping methods for Y
+    ## assign i for Y group
+    sumstats, chrom_df_y = _quick_assign_i_with_rank(sumstats, 
+                              chrpad=ychrpad, 
+                              use_rank=use_rank, 
+                              chrom=ref_chrom,
+                              pos=ref_pos,
+                              verbose=verbose)
+    
+    sumstats = sumstats.rename(columns={"i":"i_y"})
+    add_y_unique = list(sumstats["_ADD"].unique()) 
+    
+    if add_b == True:
+        sumstats["i"] = sumstats["i_x"]
+        fig,log = mqqplot(sumstats,
+                        chrom=chrom,
+                        pos=pos,
+                        p=p,
+                        mlog10p=mlog10p,
+                        snpid=snpid,
+                        scaled=scaled,
+                        log=log, 
+                        mode="b",
+                        figax=(fig,ax2),
+                        _chrom_df_for_i = chrom_df_b,
+                        _invert=False,
+                        _if_quick_qc=False,
+                        **mqq_kwargs
+                        )
+    ## 
+    #min_xy = min(min(sumstats["i_x"]),min(sumstats["i_y"]))
+    #max_xy = max(max(sumstats["i_x"]),max(sumstats["i_y"]))
+    
+    ## determine color 
+
+    ## determine dot size
+
+    ## plot  
+    legend = True
+    style=None
+    linewidth=0
+    edgecolor="black"
+
+    palette = sns.color_palette(colors,n_colors=sumstats[group].nunique()) 
+    
+    #for index,g in enumerate(sumstats[group].unique()):
+    #    
+    #    palette = sns.color_palette("dark:{}".format(colors[index]), as_cmap=True)
+    #    
+    #    plot = sns.scatterplot(data=sumstats.loc[sumstats[group]==g,:], x='i_x', y='i_y',
+    #            hue="scaled_P",
+    #            palette=palette,
+    #            size="scaled_P",
+    #            alpha=alpha,
+    #            sizes=sizes,
+    #            legend=legend,
+    #            style=style,
+    #            linewidth=linewidth,
+    #            edgecolor = edgecolor,
+    #            zorder=2,
+    #            ax=ax1)   
+    
+    plot = sns.scatterplot(data=sumstats, x='i_x', y='i_y',
+            hue=group,
+            palette=palette,
+            size="scaled_P",
+            alpha=alpha,
+            sizes=sizes,
+            legend=legend,
+            style=style,
+            linewidth=linewidth,
+            edgecolor = edgecolor,
+            zorder=2,
+            ax=ax1)   
+    
+    handles, labels = ax1.get_legend_handles_labels()
+    new_labels = []
+    ncol = len(labels)
+    for i in labels:
+        if i==group:
+            new_labels.append("Group")
+        elif i=="scaled_P":
+            new_labels.append("$-log_{10}(P)$")
+        else:
+            new_labels.append(i)
+    
+    ax1.legend(labels = new_labels,  handles=handles, loc="lower center", bbox_to_anchor=(.45, -0.17), 
+                    ncol=ncol, scatterpoints=2, title=None, frameon=False)
+
+    ## add vertical line
+    for i in add_x_unique:
+        ax1.axvline(x = i+0.5, linewidth = grid_linewidth,color=grid_linecolor,zorder=1000 )
+    for i in add_y_unique:
+        ax1.axhline(y = i+0.5,  linewidth = grid_linewidth,color=grid_linecolor,zorder=1000 )
+
+    
+    ## add X tick label 
+    ax1 = _process_xtick(ax1, chrom_df_x, xtick_chr_dict, fontsize, font_family, log=log,verbose=True)
+    ## add Y tick label 
+    ax1 = _process_ytick(ax1, chrom_df_y, ytick_chr_dict, fontsize, font_family, log=log,verbose=True)
+    
+    ## set x y lim
+    ax1.set_ylim([0.5,sumstats["i_y"].max()+1])
+    ax1.set_xlim([0.5,sumstats["i_x"].max()+1])
+
+    ## set x y label
+
+    xlabel = "pQTL position"
+    ax1.set_xlabel(xlabel,fontsize=fontsize,family=font_family)
+    ylabel = "location of the gene encoding the target protein"
+    ax1.set_ylabel(ylabel,fontsize=fontsize,family=font_family)
+    
+    save_figure(fig = fig, save = save, keyword="gwheatmap",  save_args=save_kwargs, log = log, verbose=verbose)
+
+    return fig, log
+
+################################################################################################################
+def _process_xtick(ax1, chrom_df, xtick_chr_dict, fontsize, font_family, log=Log(),verbose=True):
+    log.write(" -Processing X ticks...",verbose=verbose)
+    ax1.set_xticks(chrom_df.astype("float64"))
+    ax1.set_xticklabels(chrom_df.index.astype("Int64").map(xtick_chr_dict),fontsize=fontsize,family=font_family)    
+    return ax1
+
+def _process_ytick(ax1, chrom_df, ytick_chr_dict, fontsize, font_family, log=Log(),verbose=True):
+    log.write(" -Processing Y ticks...",verbose=verbose)
+    ax1.set_yticks(chrom_df.astype("float64"))
+    ax1.set_yticklabels(chrom_df.index.astype("Int64").map(ytick_chr_dict),fontsize=fontsize,family=font_family)    
+    return ax1

gwaslab/viz_plot_stackedregional.py

@@ -95,6 +95,10 @@ def plot_stacked_mqq(objects,
         if "family" not in title_args.keys():
             title_args["family"] = "Arial"
     # create figure and axes ##################################################################################################################
+    #
+    # subplot_height : subplot height
+    # figsize : Width, height in inches
+
     if mode=="r":
         if len(vcfs)==1:
             vcfs = vcfs *len(sumstats_list)
@@ -107,14 +111,17 @@ def plot_stacked_mqq(objects,
         else:
             height_ratios = [1 for i in range(n_plot_plus_gene_track-1)]+[gene_track_height]
         
-        fig_args["figsize"] = [16,subplot_height*n_plot_plus_gene_track]
+        if "figsize" not in fig_args.keys():
+            fig_args["figsize"] = [16,subplot_height*n_plot_plus_gene_track]
+
         fig, axes = plt.subplots(n_plot_plus_gene_track, 1, sharex=True, 
                              gridspec_kw={'height_ratios': height_ratios},
                              **fig_args)
         plt.subplots_adjust(hspace=region_hspace)
     elif mode=="m":
         n_plot = len(sumstats_list)
-        fig_args["figsize"] = [10,subplot_height*n_plot]
+        if "figsize" not in fig_args.keys():
+            fig_args["figsize"] = [10,subplot_height*n_plot]
         fig, axes = plt.subplots(n_plot, 1, sharex=True, 
                              gridspec_kw={'height_ratios': [1 for i in range(n_plot)]},
                              **fig_args)
@@ -122,8 +129,8 @@ def plot_stacked_mqq(objects,
         vcfs = [None for i in range(n_plot)]
     elif mode=="mqq":
         n_plot = len(objects)
-#
-        fig_args["figsize"] = [10,subplot_height*n_plot]
+        if "figsize" not in fig_args.keys():
+            fig_args["figsize"] = [10,subplot_height*n_plot]
         fig, axes = plt.subplots(n_plot, 2, sharex=True, 
                              gridspec_kw={'height_ratios': [1 for i in range(n_plot-1)],
                                           'width_ratios':[mqqratio,1]},