gwaslab 3.5.2__tar.gz → 3.5.4__tar.gz

{gwaslab-3.5.2/src/gwaslab.egg-info → gwaslab-3.5.4}/PKG-INFO

@@ -1,6 +1,6 @@
 Metadata-Version: 2.1
 Name: gwaslab
-Version: 3.5.2
+Version: 3.5.4
 Summary: A collection of handy tools for GWAS SumStats
 Author-email: Yunye <yunye@gwaslab.com>
 Project-URL: Homepage, https://cloufield.github.io/gwaslab/

{gwaslab-3.5.2 → gwaslab-3.5.4}/pyproject.toml

@@ -7,7 +7,7 @@ build-backend = "setuptools.build_meta"
 
 [project]
 name = "gwaslab"
-version = "3.5.2"
+version = "3.5.4"
 authors = [
   { name="Yunye", email="yunye@gwaslab.com" },
 ]

{gwaslab-3.5.2 → gwaslab-3.5.4}/src/gwaslab/g_version.py

@@ -15,8 +15,8 @@ def _get_version():
 def gwaslab_info():
     # version meta information
     dic={
-   "version":"3.5.2",
-   "release_date":"20241203"
+   "version":"3.5.4",
+   "release_date":"20241218"
     }
     return dic   
 

{gwaslab-3.5.2 → gwaslab-3.5.4}/src/gwaslab/io_process_args.py

@@ -9,14 +9,26 @@ def _extract_kwargs(prefix:str, default:dict, kwargs:dict) -> dict:
     # kwargs: all local kwargs  + args + kwargs
 
     extracted = []
+    extracted_single=dict()
     for key,value in kwargs.items():
+        # kwargs or args
         if key=="kwargs" or key=="args":
             for key_nested,value_nested in kwargs[key].items():
                 if prefix in key_nested and "arg" in key_nested:
-                    extracted.append(value_nested)
+
+                    if len(key_nested.split("_"))<3:
+                        extracted.append(value_nested)
+                    ##
+                    ## prefix_arg_fontsize
+                    else:
+                        print(key_nested.split("_")[-1], value)
+                        extracted_single[key_nested.split("_")[-1]] = value_nested
         else:
+            # local kwargs
             if prefix in key and "arg" in key:
                 extracted.append(value)
+    if len(extracted_single.keys()) >0:
+        extracted.append(extracted_single)
     merged_arg = _merge_and_sync_dic(extracted, default)
     return merged_arg
 

{gwaslab-3.5.2 → gwaslab-3.5.4}/src/gwaslab/io_to_formats.py

@@ -28,6 +28,7 @@ from gwaslab.util_in_filter_value import _extract
 def _to_format(sumstats,
               path="./sumstats",
               fmt="gwaslab",   
+              tab_fmt="tsv",
               extract=None,
               exclude=None,
               cols=None,
@@ -39,7 +40,6 @@ def _to_format(sumstats,
               n=None,
               no_status=False,
               output_log=True,
-              to_csvargs=None,
               float_formats=None,
               xymt_number=False,
               xymt=None,
@@ -47,20 +47,30 @@ def _to_format(sumstats,
               meta=None,
               ssfmeta=False,
               md5sum=False,
+              gzip=True,
               bgzip=False,
               tabix=False,
               tabix_indexargs={},
+              to_csvargs=None,
+              to_tabular_kwargs=None,
               log=Log(),
               verbose=True):
     
-    if  to_csvargs is None:
-        to_csvargs = {}
+    if to_csvargs is None:
+        to_csvargs=dict()
+    if tabix_indexargs is None:
+        tabix_indexargs=dict()
+    if to_tabular_kwargs is None:
+        to_tabular_kwargs=dict()
     if  float_formats is None:
-        float_formats={}
+        float_formats=dict()
     if cols is None:
         cols=[]
     if xymt is None:
         xymt = ["X","Y","MT"]
+    non_gzip_tab_fmt = ["parquet"]
+    non_md5sum_tab_fmt = ["parquet"]
+
     onetime_log = copy.deepcopy(log)
 
     #######################################################################################################
@@ -154,6 +164,7 @@ def _to_format(sumstats,
         tofmt(output,
               path=path,
               fmt=fmt,
+              tab_fmt=tab_fmt,
               cols=cols,
               suffix=suffix,
               build=build,
@@ -164,9 +175,13 @@ def _to_format(sumstats,
                 chr_prefix=chr_prefix,
                 meta=meta,
                 ssfmeta=ssfmeta,
+                gzip=gzip,
                 bgzip=bgzip,
+                non_gzip_tab_fmt=non_gzip_tab_fmt,
+                non_md5sum_tab_fmt=non_md5sum_tab_fmt,
                 tabix=tabix,
                 tabix_indexargs=tabix_indexargs,
+                to_tabular_kwargs=to_tabular_kwargs,
                 md5sum=md5sum,
                 xymt_number=xymt_number,
                 xymt=xymt)
@@ -186,6 +201,7 @@ def tofmt(sumstats,
           path=None,
           suffix=None,
           fmt=None,
+          tab_fmt="csv",
           cols=[],
           xymt_number=False,
           xymt=["X","Y","MT"],
@@ -194,15 +210,16 @@ def tofmt(sumstats,
           ssfmeta=False,
           md5sum=False,
           bgzip=False,
+          gzip=True,
+          non_gzip_tab_fmt=None,
+          non_md5sum_tab_fmt=None,
           tabix=False,
-          tabix_indexargs={},
+          tabix_indexargs=None,
           verbose=True,
           no_status=False,
           log=Log(),
-          to_csvargs=None):
-    
-    if to_csvargs is None:
-        to_csvargs=dict()
+          to_csvargs=None,
+          to_tabular_kwargs=None):
     
     if fmt in ["ssf"]: 
         xymt_number=True
@@ -336,36 +353,86 @@ def tofmt(sumstats,
         _bgzip_tabix_md5sum(path, fmt, bgzip, md5sum, tabix, tabix_indexargs, log, verbose)
     
     ####################################################################################################################
-    elif fmt in get_formats_list():      
+    elif fmt in get_formats_list() :      
         # tabular 
         log.write(" -"+fmt+" format will be loaded...",verbose=verbose)
         meta_data,rename_dictionary = get_format_dict(fmt,inverse=True)
         print_format_info(fmt=fmt, meta_data=meta_data,rename_dictionary=rename_dictionary,verbose=verbose, log=log, output=True)
         
-        yaml_path = path + "."+suffix+".tsv-meta.yaml"
-        path = path + "."+suffix+".tsv.gz"
+        # determine if gzip or not / create path for output
+        if gzip ==True and tab_fmt not in non_gzip_tab_fmt:
+            path = path + "."+suffix+".{}.gz".format(tab_fmt)
+        else:
+            path = path + "."+suffix+".{}".format(tab_fmt)
+
+        yaml_path = path + "."+suffix+".{}-meta.yaml".format(tab_fmt)
         log.write(" -Output path:",path, verbose=verbose) 
-        
+
         sumstats,to_csvargs = _configure_output_cols_and_args(sumstats, rename_dictionary, cols, no_status, path, meta_data, to_csvargs, log, verbose)
         
         log.write(" -Writing sumstats to: {}...".format(path),verbose=verbose)
-        try:
-            fast_to_csv(sumstats, path, to_csvargs=to_csvargs, compress=True, write_in_chunks=True)
-        except:
-            log.write(f"Error in using fast_to_csv. Falling back to original implementation.",verbose=verbose)
-            sumstats.to_csv(path, index=None, **to_csvargs)
-
-        if md5sum == True: 
-            md5_value = md5sum_file(path,log,verbose)
-        else:
-            md5_value = calculate_md5sum_file(path)
         
+        #if tab_fmt=="tsv" or tab_fmt=="csv":
+        #    try:
+        #        log.write(f"  -Fast to csv mode...",verbose=verbose)
+        #        fast_to_csv(sumstats, path, to_csvargs=to_csvargs, compress=True, write_in_chunks=True)
+        #    except:
+        #        log.write(f"Error in using fast_to_csv. Falling back to original implementation.",verbose=verbose)
+        #        sumstats.to_csv(path, index=None, **to_csvargs)
+#
+        #elif tab_fmt=="parquet":
+        #    sumstats.to_parquet(path, index=None, **to_tabular_kwargs)
+        _write_tabular(sumstats,rename_dictionary, path, tab_fmt, to_csvargs, to_tabular_kwargs, log, verbose)
+        
+        if tab_fmt not in non_md5sum_tab_fmt and "@" not in path:
+            if md5sum == True: 
+                # write a md5sum file
+                md5_value = md5sum_file(path,log,verbose)
+            else:
+                # calculate md5sum without saveing a file
+                md5_value = calculate_md5sum_file(path)
+        else:
+            md5_value = "NA"
+
         ## update ssf-style meta data and export to yaml file
         _configure_ssf_meta(sumstats, fmt, ssfmeta, meta, meta_data, path, md5_value, yaml_path, log, verbose)
         
         return sumstats  
     
 ####################################################################################################################
+def _write_tabular(sumstats,rename_dictionary, path, tab_fmt, to_csvargs, to_tabular_kwargs, log, verbose):
+    chr_header = rename_dictionary["CHR"]
+    if tab_fmt=="tsv" or tab_fmt=="csv":
+        try:
+            log.write(f"  -Fast to csv mode...",verbose=verbose)
+            if "@" in path:
+                log.write(f"  -@ detected: writing each chromosome to a single file...",verbose=verbose)
+                log.write("  -Chromosomes:{}...".format(list(sumstats["CHR"].unique())),verbose=verbose)
+                for single_chr in list(sumstats["CHR"].unique()):
+                    single_path = path.replace("@",single_chr)
+
+                    fast_to_csv(sumstats.loc[sumstats[chr_header]==single_chr,:],
+                                 single_path, 
+                                 to_csvargs=to_csvargs, compress=True, write_in_chunks=True)
+            else:
+                fast_to_csv(sumstats, path, to_csvargs=to_csvargs, compress=True, write_in_chunks=True)
+        except:
+            log.write(f"Error in using fast_to_csv. Falling back to original implementation.",verbose=verbose)
+            if "@" in path:
+                log.write(f"  -@ detected: writing each chromosome to a single file...",verbose=verbose)
+                log.write("  -Chromosomes:{}...".format(list(sumstats["CHR"].unique())),verbose=verbose)
+                for single_chr in list(sumstats["CHR"].unique()):
+                    single_path = path.replace("@",single_chr)
+
+                    sumstats.loc[sumstats[chr_header]==single_chr,:].to_csv(path, index=None, **to_csvargs)
+            else:
+                sumstats.to_csv(path, index=None, **to_csvargs)
+
+    elif tab_fmt=="parquet":
+        sumstats.to_parquet(path, index=None, **to_tabular_kwargs)
+
+
+
 def fast_to_csv(dataframe, path, to_csvargs=None, compress=True, write_in_chunks=True):
         df_numpy = dataframe.to_numpy()
 

{gwaslab-3.5.2 → gwaslab-3.5.4}/src/gwaslab/viz_aux_annotate_plot.py

@@ -31,6 +31,7 @@ def annotate_single(
     arm_scale_d,
     arm_offset,
     anno_adjust,
+    anno_xshift,
     anno_fixed_arm_length,
     maxy,
     anno_fontsize,
@@ -158,6 +159,8 @@ def annotate_single(
             if anno_fixed_arm_length is not None:
                 xytext=(row["i"],row["scaled_P"] + 0.2 + anno_fixed_arm_length)
 
+            if anno_xshift is not None:
+                xytext = (xytext[0] +(anno_xshift*y_span), xytext[1])
             ################################################################################################################################
             # if not changing the directions of some annotation arror arms 
             if anno_count not in anno_d.keys():

{gwaslab-3.5.2 → gwaslab-3.5.4}/src/gwaslab/viz_aux_chromatin.py

@@ -99,7 +99,7 @@ def _plot_chromatin_state(region_chromatin_files,
             color=color_dict_i[row["STATE_i"]]
             ax.plot([offset_i + row["START"] ,offset_i + row["END"]],
                     [i*0.1,i*0.1],
-                    c=color/255,linewidth=points_for_01,solid_capstyle="butt")
+                    c=color/255,linewidth=points_for_01,solid_capstyle="butt",rasterized=True)
     
     ## add stripe label
     if len(region_chromatin_labels) == len(region_chromatin_files):

{gwaslab-3.5.2 → gwaslab-3.5.4}/src/gwaslab/viz_plot_miamiplot2.py

@@ -168,15 +168,22 @@ def plot_miami2(
     
     if titles is None:
         titles=["",""]
+    
+    
+    titles_pad_adjusted=[1,0]
     if titles_pad is None:
         titles_pad=[0.2,0.2]
-        titles_pad_adjusted=[1 + titles_pad[0], - titles_pad[1]]
         if "anno" in mqq_args.keys():
-            titles_pad_adjusted= [1 + titles_pad[0], -titles_pad[1]]
+            if mqq_args["anno"] is not None:
+                titles_pad_adjusted= [1 + titles_pad[0], -titles_pad[1]]
         if "anno1" in mqq_args.keys():
-            titles_pad_adjusted[0]= 1 + titles_pad[0]
+            if mqq_args["anno1"] is not None:
+                titles_pad_adjusted[0]= 1 + titles_pad[0]
         if "anno2" in mqq_args.keys():
-            titles_pad_adjusted[1]=  - titles_pad[1]
+            if mqq_args["anno2"] is not None:
+                titles_pad_adjusted[1]=  - titles_pad[1]
+    else:
+        titles_pad_adjusted=[1 + titles_pad[0], -titles_pad[1]]
 
     if merged_sumstats is None:
     ## load sumstats1 ###########################################################################################################

{gwaslab-3.5.2 → gwaslab-3.5.4}/src/gwaslab/viz_plot_mqqplot.py

@@ -25,6 +25,7 @@ from gwaslab.viz_plot_qqplot import _plot_qq
 from gwaslab.hm_harmonize_sumstats import auto_check_vcf_chr_dict
 from gwaslab.viz_plot_regional2 import _plot_regional
 from gwaslab.viz_plot_regional2 import process_vcf
+from gwaslab.viz_plot_regional2 import _get_lead_id
 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
@@ -107,6 +108,7 @@ def mqqplot(insumstats,
           region_flank_factor = 0.05,
           region_anno_bbox_args = None,
           region_marker_shapes=None,
+          region_legend_marker=True,
           cbar_title='LD $r^{2}$ with variant',
           cbar_fontsize = None,
           cbar_font_family = None,
@@ -138,6 +140,7 @@ def mqqplot(insumstats,
           anno_source = "ensembl",
           anno_gtf_path=None,
           anno_adjust=False,
+          anno_xshift=None,
           anno_max_iter=100,
           arrow_kwargs=None,
           arm_offset=None,
@@ -233,7 +236,7 @@ def mqqplot(insumstats,
     if "dpi" not in fig_args.keys():
         fig_args["dpi"] = dpi
     if region_anno_bbox_args is None:
-        region_anno_bbox_args = dict()
+        region_anno_bbox_args = {"ec":"None","fc":"None"}
     if anno_set is None:
         anno_set=list()
     if anno_alias is None:
@@ -626,7 +629,8 @@ def mqqplot(insumstats,
             sumstats["chr_hue"]=sumstats["LD"]
                 
         ## default seetings
-        
+        # assign to_plot for scatter plot
+        to_plot = None
         palette = sns.color_palette(colors,n_colors=sumstats[chrom].nunique())  
 
         legend = None
@@ -639,14 +643,17 @@ def mqqplot(insumstats,
             legend=None
             linewidth=1
             if len(region_ref) == 1:
+                # hide lead variants -> add back in region plot 
                 palette = {100+i:region_ld_colors[i] for i in range(len(region_ld_colors))}
                 scatter_args["markers"]= {(i+1):m for i,m in enumerate(region_marker_shapes[:2])}
                 if region_ref[0] is None:
                     id_to_hide = sumstats["scaled_P"].idxmax()
-                    sumstats.loc[id_to_hide,"s"] = -100
+                    to_plot = sumstats.drop(id_to_hide, axis=0)
                 else:
-                    sumstats.loc[sumstats["SNPID"]==region_ref[0],"s"] = -100
-                marker_size=(0,marker_size[1])
+                    #id_to_hide = sumstats[sumstats["SNPID"]==region_ref[0],"scaled_P"].idxmax()
+                    id_to_hide = _get_lead_id(sumstats, region_ref, log=log, verbose=verbose)
+                    if id_to_hide is not None:
+                        to_plot = sumstats.drop(id_to_hide, axis=0)
                 style="SHAPE"
             else:
                 palette = {}
@@ -671,12 +678,13 @@ def mqqplot(insumstats,
                 scatter_args["markers"]= {(i+1):m for i,m in enumerate(region_marker_shapes[:len(region_ref)])}
                 style="SHAPE"
 
-            
+        
         ## if highlight 
         highlight_i = pd.DataFrame()
         if len(highlight) >0:
+            to_plot = sumstats
             log.write(" -Creating background plot...",verbose=verbose)
-            plot = sns.scatterplot(data=sumstats, x='i', y='scaled_P',
+            plot = sns.scatterplot(data=to_plot, x='i', y='scaled_P',
                                hue='chr_hue',
                                palette=palette,
                                legend=legend,
@@ -688,7 +696,7 @@ def mqqplot(insumstats,
             if pd.api.types.is_list_like(highlight[0]) and highlight_chrpos==False:
                 for i, highlight_set in enumerate(highlight):
                     log.write(" -Highlighting set {} target loci...".format(i+1),verbose=verbose)
-                    sns.scatterplot(data=sumstats.loc[sumstats["HUE"]==i], x='i', y='scaled_P',
+                    sns.scatterplot(data=to_plot.loc[to_plot["HUE"]==i], x='i', y='scaled_P',
                         hue="HUE",
                         palette={i:highlight_color[i%len(highlight_color)]},
                         legend=legend,
@@ -697,10 +705,10 @@ def mqqplot(insumstats,
                         sizes=(marker_size[0]+1,marker_size[1]+1),
                         linewidth=linewidth,
                         zorder=3+i,ax=ax1,edgecolor=edgecolor,**scatter_args)  
-                highlight_i = sumstats.loc[~sumstats["HUE"].isna(),"i"].values
+                highlight_i = to_plot.loc[~to_plot["HUE"].isna(),"i"].values
             else:
                 log.write(" -Highlighting target loci...",verbose=verbose)
-                sns.scatterplot(data=sumstats.loc[sumstats["HUE"]==0], x='i', y='scaled_P',
+                sns.scatterplot(data=to_plot.loc[to_plot["HUE"]==0], x='i', y='scaled_P',
                     hue="HUE",
                     palette={0:highlight_color},
                     legend=legend,
@@ -711,7 +719,7 @@ def mqqplot(insumstats,
                     zorder=3,ax=ax1,edgecolor=edgecolor,**scatter_args)  
                 # for annotate
                 if highlight_chrpos==False:
-                    highlight_i = sumstats.loc[sumstats[snpid].isin(highlight),"i"].values
+                    highlight_i = to_plot.loc[to_plot[snpid].isin(highlight),"i"].values
                 else:
                     highlight_i = []
         
@@ -749,7 +757,8 @@ def mqqplot(insumstats,
                 s = "s"
                 hue = 'chr_hue'
                 hue_norm=None
-                to_plot = sumstats
+                if to_plot is None:
+                    to_plot = sumstats
                 log.write(" -Creating background plot...",verbose=verbose)
                 plot = sns.scatterplot(data=to_plot, x='i', y='scaled_P',
                        hue=hue,
@@ -824,6 +833,7 @@ def mqqplot(insumstats,
                                 region_title=region_title,
                                 region_title_args=region_title_args,
                                 region_ld_legend = region_ld_legend,
+                                region_legend_marker=region_legend_marker,
                                 region_ld_threshold = region_ld_threshold,
                                 region_ld_colors = region_ld_colors,
                                 palette = palette,
@@ -1006,6 +1016,7 @@ def mqqplot(insumstats,
                                 arm_scale_d=arm_scale_d,
                                 arm_offset=arm_offset,
                                 anno_adjust=anno_adjust,
+                                anno_xshift=anno_xshift,
                                 anno_fixed_arm_length=anno_fixed_arm_length,
                                 maxy=maxy,
                                 anno_fontsize= anno_fontsize,

{gwaslab-3.5.2 → gwaslab-3.5.4}/src/gwaslab/viz_plot_regional2.py

@@ -70,6 +70,7 @@ def _plot_regional(
     palette=None,
     region_recombination = True,
     region_protein_coding=True,
+    region_legend_marker=True,
     region_flank_factor = 0.05,
     track_font_family="Arial",
     taf=[4,0,0.95,1,1],
@@ -133,6 +134,7 @@ def _plot_regional(
                             region_ref_index_dic=region_ref_index_dic,
                             region_marker_shapes=region_marker_shapes,
                             palette=palette,
+                            region_legend_marker=region_legend_marker,
                             fig=fig)
         else:
             cbar=None
@@ -244,6 +246,10 @@ def _plot_regional(
 
 # + ###########################################################################################################################################################################
 def _get_lead_id(sumstats=None, region_ref=None, log=None, verbose=True):
+    # region_ref_single (not none) -> specified variant ID
+    # convert region_ref_single -> lead_id(index)
+    
+    #
     region_ref_to_check = copy.copy(region_ref)
     try: 
         if len(region_ref_to_check)>0 and type(region_ref_to_check) is not str:
@@ -253,24 +259,30 @@ def _get_lead_id(sumstats=None, region_ref=None, log=None, verbose=True):
     
     lead_id=None
     
+    # match by rsID 
     if "rsID" in sumstats.columns:
         lead_id = sumstats.index[sumstats["rsID"] == region_ref_to_check].to_list()
-
+    # match by SNPID
     if lead_id is None and "SNPID" in sumstats.columns:
         lead_id = sumstats.index[sumstats["SNPID"] == region_ref_to_check].to_list()
 
+    # if duplicated, select the first one
     if type(lead_id) is list:
         if len(lead_id)>0:
             lead_id = int(lead_id[0])
 
+
     if region_ref_to_check is not None:
         if type(lead_id) is list:
             if len(lead_id)==0 :
                 #try:
+                # if region_ref_to_check is in CHR:POS:NEA:EA format
                 matched_snpid = re.match("(chr)?[0-9]+:[0-9]+:[ATCG]+:[ATCG]+", region_ref_to_check,  re.IGNORECASE)    
                 if matched_snpid is None:
+                    # if not, pass
                     pass
                 else:
+                    # if region_ref_to_check is in CHR:POS:NEA:EA format, match by CHR:POS:NEA:EA
                     lead_snpid = matched_snpid.group(0).split(":")
                     if len(lead_snpid)==4:
                         lead_chr= int(lead_snpid[0])
@@ -311,12 +323,13 @@ def _pinpoint_lead(sumstats,ax1,region_ref, region_ref_total_n, lead_color, mark
     else:
         lead_id = _get_lead_id(sumstats, region_ref, log, verbose)
     
-    if region_ref_total_n <2:
-        # single-ref mode
-        marker_shape = region_marker_shapes[sumstats.loc[lead_id,"SHAPE"]]
-    else:
-        # multi-ref mode
-        marker_shape = region_marker_shapes[sumstats.loc[lead_id,"SHAPE"]-1]
+    if lead_id is not None:
+        if region_ref_total_n <2:
+            # single-ref mode
+            marker_shape = region_marker_shapes[sumstats.loc[lead_id,"SHAPE"]]
+        else:
+            # multi-ref mode
+            marker_shape = region_marker_shapes[sumstats.loc[lead_id,"SHAPE"]-1]
 
     if lead_id is not None:
         ax1.scatter(sumstats.loc[lead_id,"i"],sumstats.loc[lead_id,"scaled_P"],
@@ -332,7 +345,7 @@ def _add_region_title(region_title, ax1,region_title_args):
     ax1.text(0.015,0.97, region_title, transform=ax1.transAxes, va="top", ha="left", region_ref=None, **region_title_args )
     return ax1
 
-def _add_ld_legend(sumstats, ax1, region_ld_threshold, region_ref,region_ref_index_dic,region_marker_shapes,fig, palette =None, position=1):
+def _add_ld_legend(sumstats, ax1, region_ld_threshold, region_ref,region_ref_index_dic,region_marker_shapes,fig, region_legend_marker=True,palette =None, position=1):
     
     width_pct = "11%"
     height_pct = "{}%".format( 14 + 7 * len(region_ref))
@@ -370,26 +383,30 @@ def _add_ld_legend(sumstats, ax1, region_ld_threshold, region_ref,region_ref_ind
     axins1.set_xlim(xmin,xmax)    
 
     ############### ##############plot marker ############## ##############
-    for group_index, ref in enumerate(region_ref):
-        x= -0.1
-        y= 0.1 + 0.2 * group_index
-        
-        if len(region_ref) <2:
-            # single-ref mode
-            marker = region_marker_shapes[group_index+1]
-        else:
-            # multi-ref mode
-            marker = region_marker_shapes[group_index]
-        
-        # ([x0,y0][x1,y1])
-        data_to_point =(axins1.bbox.get_points()[1][0]-axins1.bbox.get_points()[0][0]) / (xmax - xmin)
-        s =  data_to_point * 0.075
-        c =  palette[(region_ref_index_dic[region_ref[group_index]]+1)*100 + len(ld_ticks)-1]
-        axins1.scatter(x, y, s=s, marker=marker,c=c, edgecolors="black", linewidths = 1,  clip_on=False, zorder=100)
+    if region_legend_marker==True:
+        for group_index, ref in enumerate(region_ref):
+            x= -0.1
+            y= 0.1 + 0.2 * group_index
+            
+            if len(region_ref) <2:
+                # single-ref mode
+                marker = region_marker_shapes[group_index+1]
+                c =  palette[(region_ref_index_dic[region_ref[group_index]]+1)*100 + len(ld_ticks)]
+            else:
+                # multi-ref mode
+                marker = region_marker_shapes[group_index]
+                c =  palette[(region_ref_index_dic[region_ref[group_index]]+1)*100 + len(ld_ticks)-1]
+            
+            # ([x0,y0][x1,y1])
+            data_to_point =(axins1.bbox.get_points()[1][0]-axins1.bbox.get_points()[0][0]) / (xmax - xmin)
+            s =  (data_to_point * 0.15*0.11/(fig.dpi/72))**2 
+            
+            axins1.scatter(x, y, s=s, marker=marker,c=c, edgecolors="black", linewidths = 1,  clip_on=False, zorder=100)
+            axins1.tick_params(axis="y", pad=data_to_point* 0.11* 0.19/(fig.dpi/72))
     
     axins1.set_xlim(0,1)   
     axins1.set_aspect('equal', adjustable='box')
-    axins1.tick_params(axis="y", pad=np.sqrt(data_to_point * 0.11))
+    #axins1.tick_params(axis="y", pad=data_to_point* 0.11* 0.19/(fig.dpi/72))
     axins1.set_title('LD $r^{2}$ with variant',loc="center",y=-0.2)
     cbar = axins1
     return ax1, cbar
@@ -467,6 +484,7 @@ def _plot_gene_track(
     font_size_in_pixels= taf[2] * pixels_per_track
     font_size_in_points =  font_size_in_pixels * pixels_per_point
     linewidth_in_points=   pixels_per_track * pixels_per_point
+
     log.write(" -plotting gene track..", verbose=verbose)
     
     sig_gene_name = "Undefined"
@@ -479,6 +497,7 @@ def _plot_gene_track(
     sig_gene_names=[]
     sig_gene_lefts=[]
     sig_gene_rights=[]
+    log.write(" -plotting genes: {}..".format(len(uniq_gene_region)), verbose=verbose)
     for index,row in uniq_gene_region.iterrows():
 
         gene_color="#020080"
@@ -497,9 +516,9 @@ def _plot_gene_track(
                 sig_gene_lefts.append(gene_track_start_i+row["start"])
                 sig_gene_rights.append(gene_track_start_i+row["end"])
 
-            # plot gene line
-            ax3.plot((gene_track_start_i+row["start"],gene_track_start_i+row["end"]),
-                        (row["stack"]*2,row["stack"]*2),color=gene_color,linewidth=linewidth_in_points/10)
+        # plot gene line
+        ax3.plot((gene_track_start_i+row["start"],gene_track_start_i+row["end"]),
+                    (row["stack"]*2,row["stack"]*2),color=gene_color,linewidth=linewidth_in_points/10,solid_capstyle="butt")
         
         # plot gene name
         if row["end"] >= region[2]:
@@ -516,6 +535,7 @@ def _plot_gene_track(
                     y=row["stack"]*2+taf[4],s=gene_anno,ha="center",va="center",color="black",style='italic',size=font_size_in_points,family=track_font_family))
     
     # plot exons
+    log.write(" -plotting exons: {}..".format(len(exons)), verbose=verbose)
     for index,row in exons.iterrows():
         exon_color="#020080" 
         for sig_gene_name, sig_gene_left, sig_gene_right in zip(sig_gene_names,sig_gene_lefts,sig_gene_rights):
@@ -698,7 +718,8 @@ def process_vcf(sumstats,
     sumstats[final_rsq_col] = 0.0
     
     if len(region_ref)==1:
-        sumstats.loc[lead_id, final_shape_col] +=1 
+        if lead_id is not None:
+            sumstats.loc[lead_id, final_shape_col] +=1 
 
     for i in range(len(region_ref)):
         ld_single = "LD_{}".format(i)

{gwaslab-3.5.2 → gwaslab-3.5.4}/src/gwaslab/viz_plot_stackedregional.py

@@ -80,6 +80,8 @@ def plot_stacked_mqq(objects,
 
     if fig_args is None:
         fig_args = {"dpi":200}
+    if save_args is None:
+        save_args = {"dpi":400,"facecolor":"white"}
     if region_lead_grid_line is None:
         region_lead_grid_line = {"alpha":0.5,"linewidth" : 2,"linestyle":"--","color":"#FF0000"}
     if region_chromatin_files is None:
@@ -94,6 +96,19 @@ def plot_stacked_mqq(objects,
     else:
         if "family" not in title_args.keys():
             title_args["family"] = "Arial"
+
+    if save is not None:
+        if type(save) is not bool:
+            if len(save)>3:
+                if save[-3:]=="pdf" or save[-3:]=="svg":
+                    log.write(" -Adjusting options for saving as pdf/svg...",verbose=verbose)
+                    fig_args["dpi"]=72
+                    if "scatter_args" not in  mqq_args.keys():
+                        mqq_args["scatter_args"]={"rasterized":True}
+                    else:
+                        mqq_args["scatter_args"]["rasterized"] = True
+                else:
+                    fig_args["dpi"] = save_args["dpi"]
     # create figure and axes ##################################################################################################################
     #
     # subplot_height : subplot height
@@ -141,6 +156,8 @@ def plot_stacked_mqq(objects,
         region_lead_grids = [i for i in range(len(axes))]
     ##########################################################################################################################################
     mqq_args_for_each_plot = _sort_args(mqq_args, n_plot)
+
+
     ##########################################################################################################################################
     # get x axis dict
     if mode=="m" or mode=="r":
@@ -238,7 +255,6 @@ def plot_stacked_mqq(objects,
     # adjust labels
     # drop labels for each plot 
     # set a common laebl for all plots
-
     #if title_box is None:
     #    title_box = dict(boxstyle='square', facecolor='white', alpha=1.0, edgecolor="black")
     #    title_box = {}
@@ -270,8 +286,9 @@ def plot_stacked_mqq(objects,
     #else:
     if title_pos is None:
         title_pos = [0.01,0.97]
-    for index,title in enumerate(titles):
-        axes[index].text(title_pos[0], title_pos[1] , title, transform=axes[index].transAxes,ha="left", va='top',zorder=999999, **title_args)
+    if titles is not None:
+        for index,title in enumerate(titles):
+            axes[index].text(title_pos[0], title_pos[1] , title, transform=axes[index].transAxes,ha="left", va='top',zorder=999999, **title_args)
     
     ##########################################################################################################################################
     # draw the line for lead variants

{gwaslab-3.5.2 → gwaslab-3.5.4}/src/gwaslab/viz_plot_trumpetplot.py

@@ -283,6 +283,8 @@ def plottrumpet(mysumstats,
     ##################################################################################################
     size_norm = (sumstats[size].min(), sumstats[size].max())
     ## if highlight  ##################################################################################################
+
+    log.write(" -Creating scatter plot...", verbose=verbose)
     dots = sns.scatterplot(data=sumstats,
                     x=maf,
                     y=beta,
@@ -295,7 +297,7 @@ def plottrumpet(mysumstats,
                     alpha=0.8,
                     zorder=2,
                     **scatter_args)
-    
+    log.write(" -Finished screating scatter plot...", verbose=verbose)
     if len(highlight) >0:
         
         legend = None
@@ -380,15 +382,13 @@ def plottrumpet(mysumstats,
     ####################################################################################################################
     
     #second_legend = ax.legend(title="Power", loc="upper right",fontsize =fontsize,title_fontsize=fontsize)
-    
+    log.write(" -Creating legends...")
     h,l = ax.get_legend_handles_labels()
     if len(ts)>0:
         l1 = ax.legend(h[:int(len(ts))],l[:int(len(ts))], title="Power", loc="upper right",fontsize =fontsize,title_fontsize=fontsize)
         for line in l1.get_lines():
             line.set_linewidth(5.0)
     if hue is None:
-        l2 = ax.legend(h[int(len(ts)):],l[int(len(ts)):], title=size, loc="lower right",fontsize =fontsize,title_fontsize=fontsize)
-    else:
         l2 = ax.legend(h[int(len(ts)):],l[int(len(ts)):], title=None, loc="lower right",fontsize =fontsize,title_fontsize=fontsize)
     if len(ts)>0:
         ax.add_artist(l1)

{gwaslab-3.5.2 → gwaslab-3.5.4/src/gwaslab.egg-info}/PKG-INFO

@@ -1,6 +1,6 @@
 Metadata-Version: 2.1
 Name: gwaslab
-Version: 3.5.2
+Version: 3.5.4
 Summary: A collection of handy tools for GWAS SumStats
 Author-email: Yunye <yunye@gwaslab.com>
 Project-URL: Homepage, https://cloufield.github.io/gwaslab/