gwaslab 3.4.47__tar.gz → 3.4.48__tar.gz

{gwaslab-3.4.47/src/gwaslab.egg-info → gwaslab-3.4.48}/PKG-INFO

@@ -1,6 +1,6 @@
 Metadata-Version: 2.1
 Name: gwaslab
-Version: 3.4.47
+Version: 3.4.48
 Summary: A collection of handy tools for GWAS SumStats
 Author-email: Yunye <yunye@gwaslab.com>
 Project-URL: Homepage, https://cloufield.github.io/gwaslab/
@@ -51,7 +51,7 @@ Warning: Known issues of GWASLab are summarized in [https://cloufield.github.io/
 ### install via pip
 
 ```
-pip install gwaslab==3.4.45
+pip install gwaslab==3.4.46
 ```
 
 ```python

{gwaslab-3.4.47 → gwaslab-3.4.48}/README.md

@@ -23,7 +23,7 @@ Warning: Known issues of GWASLab are summarized in [https://cloufield.github.io/
 ### install via pip
 
 ```
-pip install gwaslab==3.4.45
+pip install gwaslab==3.4.46
 ```
 
 ```python

{gwaslab-3.4.47 → gwaslab-3.4.48}/pyproject.toml

@@ -7,7 +7,7 @@ build-backend = "setuptools.build_meta"
 
 [project]
 name = "gwaslab"
-version = "3.4.47"
+version = "3.4.48"
 authors = [
   { name="Yunye", email="yunye@gwaslab.com" },
 ]

{gwaslab-3.4.47 → gwaslab-3.4.48}/src/gwaslab/g_Sumstats.py

@@ -76,6 +76,8 @@ from gwaslab.util_ex_ldsc import _estimate_rg_by_ldsc
 from gwaslab.util_ex_ldsc import _estimate_h2_cts_by_ldsc
 from gwaslab.util_ex_ldsc import _estimate_partitioned_h2_by_ldsc 
 from gwaslab.bd_get_hapmap3 import gethapmap3
+from gwaslab.util_abf_finemapping import abf_finemapping
+from gwaslab.util_abf_finemapping import make_cs
 import gc
 
 #20220309
@@ -135,6 +137,7 @@ class Sumstats():
         self.data = pd.DataFrame()
         self.log = Log()
         self.ldsc_h2 = None
+        self.ldsc_h2_results = None
         self.ldsc_rg = None
         self.ldsc_h2_cts = None
         self.ldsc_partitioned_h2_summary = None
@@ -756,13 +759,20 @@ class Sumstats():
         else:
             output = lambdaGC(self.data[["CHR",mode]],mode=mode,**kwargs)
             self.meta["Genomic inflation factor"] = output
-            return output 
+            return output
+        
+    def abf_finemapping(self, region=None, chrpos=None, snpid=None,**kwargs):        
+        region_data = abf_finemapping(self.data.copy(),region=region,chrpos=chrpos,snpid=snpid,log=self.log, **kwargs)
+        credible_sets = make_cs(region_data,threshold=0.95,log=self.log)
+        return region_data, credible_sets
+    
+    
 ## LDSC ##############################################################################################
     def estimate_h2_by_ldsc(self, build=None, verbose=True, match_allele=True, **kwargs):
         if build is None:
             build = self.meta["gwaslab"]["genome_build"]
         insumstats = gethapmap3(self.data.copy(), build=build, verbose=verbose , match_allele=True, how="right" )
-        self.ldsc_h2 = _estimate_h2_by_ldsc(insumstats=insumstats, log=self.log, verbose=verbose, **kwargs)
+        self.ldsc_h2, self.ldsc_h2_results = _estimate_h2_by_ldsc(insumstats=insumstats, log=self.log, verbose=verbose, **kwargs)
     
     def estimate_rg_by_ldsc(self, build=None, verbose=True, match_allele=True, **kwargs):
         if build is None:

{gwaslab-3.4.47 → gwaslab-3.4.48}/src/gwaslab/g_version.py

@@ -15,8 +15,8 @@ def _get_version():
 def gwaslab_info():
     # version meta information
     dic={
-   "version":"3.4.47",
-   "release_date":"20240703"
+   "version":"3.4.48",
+   "release_date":"20240822"
     }
     return dic   
 

{gwaslab-3.4.47 → gwaslab-3.4.48}/src/gwaslab/io_read_ldsc.py

@@ -198,16 +198,29 @@ def read_greml(filelist=[]):
     return summary
 
 def parse_ldsc_summary(ldsc_summary):
-    summary = pd.DataFrame(columns = ['h2_obs', 'h2_se','Lambda_gc','Mean_chi2','Intercept','Intercept_se',"Ratio","Ratio_se"])
+    
     lines = ldsc_summary.split("\n")
+    
+    columns = ['h2_obs', 'h2_se','Lambda_gc','Mean_chi2','Intercept','Intercept_se',"Ratio","Ratio_se","Catagories"]
+    
+    summary = pd.DataFrame(columns = columns)
+    
     row={}
+    
     try:
         objects = re.compile('[a-zA-Z\s\d]+:|[-0-9.]+[e]?[-0-9.]+|NA').findall(lines[0])
         row["h2_obs"]=objects[1]
         row["h2_se"]=objects[2]
 
-        ##next line lambda gc
+        ##check categories
+        if len(lines) == 6:
+            objects = re.compile('  -Categories:(.+)').findall(lines[1])
+            row["Catagories"] = objects[0].strip()
+            lines.pop(1)
+        else:
+            row["Catagories"] = "NA"
 
+        ##next line lambda gc
         objects = re.compile('[a-zA-Z\s\d]+:|[-0-9.]+[e]?[-0-9.]+|NA').findall(lines[1])
         row["Lambda_gc"] = objects[1]
         ##next line Mean_chi2
@@ -240,6 +253,7 @@ def parse_ldsc_summary(ldsc_summary):
         row["Intercept_se"]="NA"
         row["Ratio"]="NA"
         row["Ratio_se"]="NA"
+        row["Catagories"] = "NA"
 
     #summary = summary.append(row,ignore_index=True)
     row = pd.DataFrame([row], columns = summary.columns)

{gwaslab-3.4.47 → gwaslab-3.4.48}/src/gwaslab/qc_fix_sumstats.py

@@ -1497,7 +1497,11 @@ def liftover_variant(sumstats,
              status="STATUS",
              from_build="19", 
              to_build="38"):
-    converter = get_lifter("hg"+from_build,"hg"+to_build)
+    try:
+        converter = get_lifter("hg"+from_build,"hg"+to_build,one_based=True)
+    except:
+        converter = get_lifter("hg"+from_build,"hg"+to_build)
+
     dic= get_number_to_chr(in_chr=False,xymt=["X","Y","M"])
     dic2= get_chr_to_number(out_chr=False)
     for i in sumstats[chrom].unique():
@@ -1549,7 +1553,6 @@ def parallelizeliftovervariant(sumstats,n_cores=1,chrom="CHR", pos="POS", from_b
         pool.close()
         pool.join()
     ############################################################################
-   
     unmap_num = len(sumstats.loc[sumstats[pos].isna(),:])    
     
     if remove is True:

gwaslab-3.4.48/src/gwaslab/util_abf_finemapping.py

@@ -0,0 +1,67 @@
+import pandas as pd
+import numpy as np
+from gwaslab.g_Log import Log
+from gwaslab.util_in_filter_value import _get_flanking_by_chrpos
+from gwaslab.util_in_filter_value import _get_flanking_by_id
+
+# Calculate PIP based on approximate Bayesian factor (ABF)
+# Wakefield, J. A bayesian measure of the probability of false discovery in genetic epidemiology studies. Am J Hum Genet 81, 208–227 (2007).
+
+
+def calc_abf(insumstats,w=0.2,log=Log(),verbose=True,**kwargs):
+
+
+
+    log.write("Start to calculate approximate Bayesian factor for {} variants".format(len(insumstats)),verbose=verbose)
+    log.write(" - Reference: akefield, J. A bayesian measure of the probability of false discovery in genetic epidemiology studies. Am J Hum Genet 81, 208–227 (2007).",verbose=verbose)
+    log.write(" - Priors for the standard deviation W of the effect size parameter β : {} ".format(w),verbose=verbose)
+    # binary -> w=0.2
+    # quant  -> w=0.15
+    omega = w**2
+    se = insumstats["SE"]
+    v = se**2
+    r = omega / (omega+v)
+    beta = insumstats["BETA"]
+    z = beta/se
+    insumstats = insumstats.copy()
+
+    # (6) ABF -> reciprocal
+    insumstats.loc[:, "log_ABF"] = 1/2* (np.log(1-r) + (r * z**2))
+    
+    return insumstats
+
+def calc_PIP(insumstats,log=Log(),verbose=True,**kwargs):
+    # Calculate the logarithmic sum of each ABF to find the logarithm of total_abf
+    log_total_abf = np.log(np.sum(np.exp(insumstats["log_ABF"] - np.max(insumstats["log_ABF"])))) + np.max(insumstats["log_ABF"])
+    insumstats = insumstats.copy()
+    log.write("Start to calculate PIP for {} variants".format(len(insumstats)),verbose=verbose)
+    # Calculate PIP on a logarithmic scale by subtracting log_total_abf from each log_abf
+    insumstats.loc[:, "log_PIP"] = insumstats['log_ABF'] - log_total_abf
+    # Convert PIP on logarithmic scale to exponential and back to normal scale
+    insumstats.loc[:, "PIP"] = np.exp(insumstats['log_PIP'])
+    return insumstats
+
+def abf_finemapping(insumstats,region=None,chrpos=None,snpid=None, log=Log(),**kwargs):
+
+    if region is not None:
+        region_data = insumstats[(insumstats["CHR"] == region[0]) & (insumstats["POS"] >= region[1]) & (insumstats["POS"] <= region[2])]
+    elif chrpos is not None:
+        region_data = _get_flanking_by_chrpos(insumstats, chrpos=chrpos,**kwargs)
+    elif snpid is not None:
+        region_data = _get_flanking_by_id(insumstats, snpid=snpid,**kwargs)
+
+    region_data = calc_abf(region_data,log=log,**kwargs)
+    region_data = calc_PIP(region_data,log=log,**kwargs)
+    return region_data
+
+def make_cs(insumstats,threshold=0.95,log=Log(),verbose=True):
+    insumstats = insumstats.sort_values(by="PIP",ascending=False)
+    pip_sum = 0
+    cs = pd.DataFrame()
+    for index, row in insumstats.iterrows():
+        cs = pd.concat([cs,pd.DataFrame(row).T])
+        pip_sum += row["PIP"]
+        if pip_sum > threshold:
+            break
+    log.write("Finished constructing a {}% credible set with {} variant(s)".format(str(threshold * 100),str(len(cs))),verbose=verbose)
+    return cs

{gwaslab-3.4.47 → gwaslab-3.4.48}/src/gwaslab/util_ex_ldsc.py

@@ -304,9 +304,16 @@ def _estimate_h2_by_ldsc(insumstats, log, verbose=True, munge=False, munge_args=
     log.write(" -LDSC log:", verbose=verbose)
     summary = estimate_h2(sumstats, default_args, log)
     
+    results_table = None
+    if type(summary) is tuple:
+        results_table = summary[1]
+        summary = summary[0]
+        log.write(" -Coefficient results have been stored in .ldsc_h2_results", verbose=verbose)
+        
+
     log.write(" -Results have been stored in .ldsc_h2", verbose=verbose)
     finished(log=log,verbose=verbose,end_line=_end_line)
-    return parse_ldsc_summary(summary)
+    return parse_ldsc_summary(summary), results_table
 
 
 ####################################################################################################################

{gwaslab-3.4.47 → gwaslab-3.4.48}/src/gwaslab/util_ex_run_clumping.py

@@ -11,7 +11,7 @@ from gwaslab.g_version import _checking_plink_version
 def _clump(insumstats, vcf=None, scaled=False, out="clumping_plink2", 
            p="P",mlog10p="MLOG10P", overwrite=False, study=None, bfile=None, 
            n_cores=1, memory=None, chrom=None, clump_p1=5e-8, clump_p2=5e-8, clump_r2=0.01, clump_kb=250,
-           log=Log(),verbose=True):
+           log=Log(),verbose=True,plink="plink",plink2="plink2"):
     ##start function with col checking##########################################################
     _start_line = "perfrom clumping"
     _end_line = "clumping"
@@ -111,7 +111,7 @@ def _clump(insumstats, vcf=None, scaled=False, out="clumping_plink2",
             bfile_to_use = bfile
 
         log.write(" -Performing clumping for CHR {}...".format(i),verbose=verbose)
-        log = _checking_plink_version(v=2, log=log)
+        log = _checking_plink_version(plink2=plink2, log=log)
         if memory is not None:
             memory_flag = "--memory {}".format(memory)
         
@@ -123,7 +123,7 @@ def _clump(insumstats, vcf=None, scaled=False, out="clumping_plink2",
         if scaled == True:
             # clumping using LOG10P
             script = """
-            plink2 \
+            {} \
                 {}\
                 --chr {} \
                 --clump {} \
@@ -136,11 +136,11 @@ def _clump(insumstats, vcf=None, scaled=False, out="clumping_plink2",
                 --clump-kb {} \
                 --threads {} {}\
                 --out {}
-            """.format(file_flag, chrom, clump, mlog10p,clump_log10_p1, clump_log10_p2, clump_r2, clump_kb, n_cores, memory_flag if memory is not None else "", out_single_chr)    
+            """.format(plink2, file_flag, chrom, clump, mlog10p,clump_log10_p1, clump_log10_p2, clump_r2, clump_kb, n_cores, memory_flag if memory is not None else "", out_single_chr)    
         else:
             # clumping using P
             script = """
-            plink2 \
+            {} \
                 {}\
                 --chr {} \
                 --clump {} \
@@ -152,7 +152,7 @@ def _clump(insumstats, vcf=None, scaled=False, out="clumping_plink2",
                 --clump-kb {} \
                 --threads {} {}\
                 --out {}
-            """.format(file_flag, chrom, clump, p, clump_p1, clump_p2, clump_r2, clump_kb, n_cores,memory_flag if memory is not None else "", out_single_chr)
+            """.format(plink2,file_flag, chrom, clump, p, clump_p1, clump_p2, clump_r2, clump_kb, n_cores,memory_flag if memory is not None else "", out_single_chr)
         
         try:
             output = subprocess.check_output(script, stderr=subprocess.STDOUT, shell=True,text=True)

{gwaslab-3.4.47 → gwaslab-3.4.48}/src/gwaslab/viz_aux_annotate_plot.py

@@ -38,6 +38,7 @@ def annotate_single(
     region,
     region_anno_bbox_args,
     skip,
+    anno_height=1,
     amode="int",
     snpid="SNPID",
     chrom="CHR",
@@ -131,7 +132,7 @@ def annotate_single(
             
             #xy=(row["i"],row["scaled_P"]+0.2)
             xy=(row["i"],row["scaled_P"]+0.01*maxy)
-            xytext=(last_pos,1.15*maxy*arm_scale)
+            xytext=(last_pos,1.15*maxy*arm_scale*anno_height)
             
             if anno_fixed_arm_length is not None:
                 armB_length_in_point = anno_fixed_arm_length

{gwaslab-3.4.47 → gwaslab-3.4.48}/src/gwaslab/viz_aux_quickfix.py

@@ -286,8 +286,9 @@ def _cut(series, mode,cutfactor,cut,skip, ylabels, cut_log, verbose, lines_to_pl
     log.write(" -Converting data above cut line...",verbose=verbose)
     if ylabels is not None:
         ylabels = pd.Series(ylabels)
-    maxy = series.max()
     series = series.copy()
+    
+    maxy = series.max()
     if "b" not in mode:
         log.write(" -Maximum -log10(P) value is "+str(maxy) +" .", verbose=verbose)
     elif "b" in mode:

{gwaslab-3.4.47 → gwaslab-3.4.48}/src/gwaslab/viz_plot_compare_effect.py

@@ -77,8 +77,10 @@ def compare_effect(path1,
         scaled2 = True
     if is_q_mc=="fdr" or is_q_mc=="bon":
         is_q = True
-    else:
-        raise ValueError("Please select either fdr or bon for is_q_mc.")
+    
+    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.")
     if save_args is None:
         save_args = {"dpi":300,"facecolor":"white"}
     if reg_box is None:

{gwaslab-3.4.47 → gwaslab-3.4.48}/src/gwaslab/viz_plot_miamiplot2.py

@@ -247,7 +247,10 @@ def plot_miami2(
         plt.subplots_adjust(hspace=region_hspace)    
     else:
         fig, ax1, ax5 = figax
-    
+
+    #if same_ylim==True:
+        #maxy = merged_sumstats[["scaled_P_1","scaled_P_2"]].max().max()
+
     log.write("Start to create Manhattan plot for sumstats1...", verbose=verbose)
     fig,log = mqqplot(merged_sumstats,
                       chrom="CHR",
@@ -284,14 +287,8 @@ def plot_miami2(
                       _if_quick_qc=False,
                      **mqq_args2)
     log.write("Finished creating Manhattan plot for sumstats2".format(_get_version()), verbose=verbose)
+    
 
-    if same_ylim==True:
-        ylim1_converted = ax1.get_ylim()
-        ylim2_converted = ax5.get_ylim()
-        if ylim1_converted > ylim2_converted:
-            ax5.set_ylim(ylim1_converted)
-        else:
-            ax1.set_ylim(ylim2_converted)
     #####################################################################################################################    
     
     ax5.set_xlabel("")

{gwaslab-3.4.47 → gwaslab-3.4.48}/src/gwaslab/viz_plot_mqqplot.py

@@ -141,6 +141,7 @@ def mqqplot(insumstats,
           anno_max_iter=100,
           arm_offset=50,
           arm_scale=1,
+          anno_height=1,
           arm_scale_d=None,
           cut=0,
           skip=0,
@@ -180,6 +181,7 @@ def mqqplot(insumstats,
           xpad=None,
           xpadl=None,
           xpadr=None,
+          xtight=False,
           chrpad=0.03, 
           drop_chr_start=False,
           title =None,
@@ -552,7 +554,8 @@ def mqqplot(insumstats,
                                                                         cut_log = cut_log,
                                                                         verbose =verbose, 
                                                                         lines_to_plot=lines_to_plot,
-                                                                        log = log)
+                                                                        log = log
+                                                                        )
     except:
         log.warning("No valid data! Please check the input.")
         return None
@@ -596,19 +599,23 @@ def mqqplot(insumstats,
             sumstats.loc[sumstats["scaled_P"]>-np.log10(sig_level_plot),"s"]=4
         sumstats["chr_hue"]=sumstats[chrom].astype("string")
 
-        if vcf_path is not None:
+        if "r" in mode:
+            if vcf_path is None:
+                sumstats["LD"]=100
+                sumstats["SHAPE"]=1
             sumstats["chr_hue"]=sumstats["LD"]
+                
         ## default seetings
         
         palette = sns.color_palette(colors,n_colors=sumstats[chrom].nunique())  
-        
 
         legend = None
         style=None
         linewidth=0
         edgecolor="black"
         # if regional plot assign colors
-        if vcf_path is not None:
+        if "r" in mode:
+            #if vcf_path is not None: 
             legend=None
             linewidth=1
             if len(region_ref) == 1:
@@ -631,10 +638,9 @@ def mqqplot(insumstats,
                         palette[(i+1)*100 + j ] = hex_color
 
                 edgecolor="none"
-                scatter_args["markers"]= region_marker_shapes[:len(region_ref)]
+                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()
@@ -977,6 +983,7 @@ def mqqplot(insumstats,
                                 region=region,
                                 region_anno_bbox_args=region_anno_bbox_args,
                                 skip=skip,
+                                anno_height=anno_height,
                                 snpid=snpid,
                                 chrom=chrom,
                                 pos=pos,
@@ -1040,7 +1047,7 @@ def mqqplot(insumstats,
         if "qq" in mode:
             ax2.set_ylim(ylim)
     
-    ax1 = _add_pad_to_x_axis(ax1, xpad, xpadl, xpadr, sumstats)
+    ax1 = _add_pad_to_x_axis(ax1, xpad, xpadl, xpadr, sumstats, pos, chrpad, xtight, log = log, verbose=verbose)
 
     # Titles 
     if title and anno and len(to_annotate)>0:
@@ -1065,20 +1072,34 @@ def mqqplot(insumstats,
 
 
 
-def _add_pad_to_x_axis(ax1, xpad, xpadl, xpadr, sumstats):
+def _add_pad_to_x_axis(ax1, xpad, xpadl, xpadr, sumstats, pos, chrpad, xtight, log, verbose):
     
-    if ax1 is not None:
-        xmin, xmax = ax1.get_xlim() 
-        
-        if xpad is not None:
-            pad = xpad* sumstats["i"].max()
-            ax1.set_xlim([xmin - pad, xmin + pad])
-        if xpadl is not None:
-            pad = xpadl* sumstats["i"].max()
-            ax1.set_xlim([xmin - pad,xmax])
-        if xpadr is not None:
-            pad = xpadr* sumstats["i"].max()
-            ax1.set_xlim([xmin, xmax + pad])
+    if xtight==True:
+        log.write(" -Adjusting X padding on both side : tight mode", verbose=verbose)
+        xmax = sumstats["i"].max()
+        xmin=  sumstats["i"].min()
+        ax1.set_xlim([xmin, xmax])
+    
+    else:
+        chrpad_to_remove = sumstats[pos].max()*chrpad
+        if ax1 is not None:
+            xmin, xmax = ax1.get_xlim() 
+            length = xmax - xmin
+            
+            if xpad is not None:
+                log.write(" -Adjusting X padding on both side: {}".format(xpad), verbose=verbose)
+                pad = xpad* length #sumstats["i"].max()
+                ax1.set_xlim([xmin - pad + chrpad_to_remove, xmax + pad - chrpad_to_remove])
+            if xpad is None and xpadl is not None:
+                log.write(" -Adjusting X padding on left side: {}".format(xpadl), verbose=verbose)
+                xmin, xmax = ax1.get_xlim() 
+                pad = xpadl*length # sumstats["i"].max()
+                ax1.set_xlim([xmin - pad + chrpad_to_remove ,xmax])
+            if xpad is None and xpadr is not None:
+                log.write(" -Adjusting X padding on right side: {}".format(xpadr), verbose=verbose)
+                xmin, xmax = ax1.get_xlim() 
+                pad = xpadr*length # sumstats["i"].max()
+                ax1.set_xlim([xmin, xmax + pad - chrpad_to_remove])
 
     return ax1
 

{gwaslab-3.4.47 → gwaslab-3.4.48}/src/gwaslab/viz_plot_regional2.py

@@ -5,6 +5,7 @@ import matplotlib.patches as patches
 import seaborn as sns
 import numpy as np
 import copy
+import re
 import scipy as sp
 from pyensembl import EnsemblRelease
 from allel import GenotypeArray
@@ -96,19 +97,20 @@ def _plot_regional(
                                         marker_size= marker_size,
                                         region_marker_shapes=region_marker_shapes,
                                         log=log,verbose=verbose)
-            if lead_id_single is not None:
-                lead_ids.append(lead_id_single)       
+            #if lead_id_single is not None:
+            lead_ids.append(lead_id_single)       
         
         # update region_ref to variant rsID or variantID / skip NAs
         new_region_ref = []
         for i in range(len(lead_ids)):
             if lead_ids[i] is None:
+                new_region_ref.append(region_ref[i])
                 continue
             if region_ref[i] is None:
-                if "rsID" in sumstats.columns:
-                    new_name = sumstats.loc[lead_ids[i],"rsID"]
-                elif "SNPID" in sumstats.columns:
+                if "SNPID" in sumstats.columns:
                     new_name = sumstats.loc[lead_ids[i],"SNPID"]
+                elif "rsID" in sumstats.columns:
+                    new_name = sumstats.loc[lead_ids[i],"rsID"]
                 else:
                     new_name = "chr{}:{}".format(sumstats.loc[lead_ids[i],"CHR"] , sumstats.loc[lead_ids[i],"POS"])
                 new_region_ref.append(new_name)
@@ -162,7 +164,6 @@ def _plot_regional(
         lead_snp_ys = []
         lead_snp_is = []
         lead_snp_is_colors = []
-        
         for i,lead_id_single in enumerate(lead_ids):
             if lead_id_single is not None:
                 lead_snp_ys.append(sumstats.loc[lead_id_single,"scaled_P"] )
@@ -258,11 +259,35 @@ def _get_lead_id(sumstats=None, region_ref=None, log=None, verbose=True):
     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 :
-                log.warning("{} not found.. Skipping..".format(region_ref_to_check))
+                #try:
+                matched_snpid = re.match("(chr)?[0-9]+:[0-9]+:[ATCG]+:[ATCG]+", region_ref_to_check,  re.IGNORECASE)    
+                if matched_snpid is None:
+                    pass
+                else:
+                    lead_snpid = matched_snpid.group(0).split(":")
+                    if len(lead_snpid)==4:
+                        lead_chr= int(lead_snpid[0])
+                        lead_pos= int(lead_snpid[1])
+                        lead_ea= lead_snpid[2]
+                        lead_nea= lead_snpid[3]
+                        chrpos_match = (sumstats["CHR"] == lead_chr) & (sumstats["POS"] == lead_pos)
+                        eanea_match = ((sumstats["EA"] == lead_ea) & (sumstats["NEA"] == lead_nea)) | ((sumstats["EA"] == lead_nea) & (sumstats["NEA"] == lead_ea)) 
+                        if "rsID" in sumstats.columns:
+                            lead_id = sumstats.index[chrpos_match&eanea_match].to_list()
+                        if "SNPID" in sumstats.columns:
+                            lead_id = sumstats.index[chrpos_match&eanea_match].to_list()     
+                if type(lead_id) is list:
+                    if len(lead_id)>0:
+                        lead_id = int(lead_id[0])   
+                        log.warning("Trying matching variant {} using CHR:POS:EA:NEA to {}... ".format(region_ref_to_check,lead_id))
+
+        if type(lead_id) is list:
+            if len(lead_id)==0 :
+                log.warning("Extracting variant: {} not found in sumstats.. Skipping..".format(region_ref_to_check))
                 #lead_id = sumstats["scaled_P"].idxmax()
                 lead_id = None
                 return lead_id
@@ -275,7 +300,7 @@ def _get_lead_id(sumstats=None, region_ref=None, log=None, verbose=True):
 
     return lead_id
 
-def _pinpoint_lead(sumstats,ax1,region_ref, lead_color, marker_size, log, verbose,region_marker_shapes):
+def _pinpoint_lead(sumstats,ax1,region_ref, lead_color, marker_size, log, verbose, region_marker_shapes):
     
     if region_ref is None:
         log.write(" -Extracting lead variant..." , verbose=verbose)
@@ -416,6 +441,11 @@ def _plot_gene_track(
     texts_to_adjust_left = []
     texts_to_adjust_middle = []
     texts_to_adjust_right = []
+
+    
+    sig_gene_names=[]
+    sig_gene_lefts=[]
+    sig_gene_rights=[]
     for index,row in uniq_gene_region.iterrows():
 
         gene_color="#020080"
@@ -426,21 +456,18 @@ def _plot_gene_track(
             gene_anno = "<-" + row["name"] 
         
 
-        sig_gene_names=[]
-        sig_gene_lefts=[]
-        sig_gene_rights=[]
+
         for lead_snp_i in lead_snp_is:
             if region_lead_grid is True and lead_snp_i > gene_track_start_i+row["start"] and lead_snp_i < gene_track_start_i+row["end"] :
-                    gene_color=region_lead_grid_line["color"]
-                    sig_gene_names.append(row["name"])
-                    sig_gene_lefts.append(gene_track_start_i+row["start"])
-                    sig_gene_rights.append(gene_track_start_i+row["end"])
+                gene_color=region_lead_grid_line["color"]
+                sig_gene_names.append(row["name"])
+                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 name
         if row["end"] >= region[2]:
             #right side
@@ -459,6 +486,7 @@ def _plot_gene_track(
     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):
+            
             if not pd.isnull(row["name"]):
                 if (region_lead_grid is True) and row["name"]==sig_gene_name:
                     exon_color = region_lead_grid_line["color"]  
@@ -468,7 +496,7 @@ def _plot_gene_track(
                 exon_color = region_lead_grid_line["color"]  
             else:
                 exon_color="#020080"
-
+        
         ax3.plot((gene_track_start_i+row["start"],gene_track_start_i+row["end"]),
                     (row["stack"]*2,row["stack"]*2),linewidth=linewidth_in_points*taf[3],color=exon_color,solid_capstyle="butt")
 
@@ -550,24 +578,42 @@ def process_vcf(sumstats,
             # figure out lead variant
             lead_id = _get_lead_id(sumstats, region_ref_single, log, verbose)
         
-        if lead_id is None:
-            sumstats[rsq] = None
-            sumstats[rsq] = sumstats[rsq].astype("float")
-            sumstats[ld_single] = 0             
-            continue
 
-        lead_pos = sumstats.loc[lead_id,pos]
+        lead_series = None
+        if lead_id is None:
+            
+            matched_snpid = re.match("(chr)?[0-9]+:[0-9]+:[ATCG]+:[ATCG]+",region_ref_single,  re.IGNORECASE)
+            
+            if matched_snpid is None:
+                sumstats[rsq] = None
+                sumstats[rsq] = sumstats[rsq].astype("float")
+                sumstats[ld_single] = 0    
+                continue    
+            else:
+                
+                lead_snpid = matched_snpid.group(0).split(":")[1:]
+                lead_pos = int(lead_snpid[0])
+                lead_snpid[0]= int(lead_snpid[0])
+                lead_series = pd.Series(lead_snpid)
+        else:
+            lead_pos = sumstats.loc[lead_id,pos]
 
+        
         # if lead pos is available: 
         if lead_pos in ref_genotype["variants/POS"]:
             
             # get ref index for lead snp
-            lead_snp_ref_index = match_varaint(sumstats.loc[lead_id,[pos,nea,ea]])
-            #lead_snp_ref_index = np.where(ref_genotype["variants/POS"] == lead_pos)[0][0]
+            if lead_series is None:
+                lead_snp_ref_index = match_varaint(sumstats.loc[lead_id,[pos,nea,ea]])
+                #lead_snp_ref_index = np.where(ref_genotype["variants/POS"] == lead_pos)[0][0]
+            else:
+                log.warning("Computing LD: {} not found in sumstats but found in reference...Still Computing...".format(region_ref_single))
+                lead_snp_ref_index = match_varaint(lead_series)
 
             # non-na other snp index
             other_snps_ref_index = sumstats["REFINDEX"].dropna().astype("int").values
             # get genotype 
+            
             lead_snp_genotype = GenotypeArray([ref_genotype["calldata/GT"][lead_snp_ref_index]]).to_n_alt()
             try:
                 if len(set(lead_snp_genotype[0]))==1:
@@ -604,10 +650,10 @@ def process_vcf(sumstats,
                 sumstats.loc[to_change_color,ld_single] = 1
             to_change_color = sumstats[rsq]>ld_threshold
             sumstats.loc[to_change_color,ld_single] = index+2
-
-        sumstats.loc[lead_id,ld_single] = len(region_ld_threshold)+2
         
-        sumstats.loc[lead_id,lead] = 1
+        if lead_series is None:
+            sumstats.loc[lead_id,ld_single] = len(region_ld_threshold)+2
+            sumstats.loc[lead_id,lead] = 1
 
     ####################################################################################################    
     final_shape_col = "SHAPE"

{gwaslab-3.4.47 → gwaslab-3.4.48}/src/gwaslab/viz_plot_stackedregional.py

@@ -57,7 +57,9 @@ def plot_stacked_mqq(objects,
                         fig_args=None,
                         region_hspace=0.05,
                         subplot_height=4,
+                        region_lead_grids = None,
                         region_lead_grid_line=None,
+                        region_ld_legends = None,
                         fontsize=9,
                         font_family="Arial",
                         build="99", 
@@ -85,6 +87,8 @@ def plot_stacked_mqq(objects,
         region_chromatin_height = len(region_chromatin_files) * region_chromatin_height
     if region_chromatin_labels is None:
         region_chromatin_labels = []
+    if region_ld_legends is None:
+        region_ld_legends = [0]
     if title_args is None:
         title_args = {"family":"Arial"}
     else:
@@ -125,7 +129,9 @@ def plot_stacked_mqq(objects,
                                           'width_ratios':[mqqratio,1]},
                                           **fig_args)
         plt.subplots_adjust(hspace=region_hspace)
-
+    
+    if region_lead_grids is None:
+        region_lead_grids = [i for i in range(len(axes))]
     ##########################################################################################################################################
     mqq_args_for_each_plot = _sort_args(mqq_args, n_plot)
     ##########################################################################################################################################
@@ -149,6 +155,10 @@ def plot_stacked_mqq(objects,
             figax = (fig,axes[index],axes[-1])
         elif mode=="mqq":
             figax = (fig,axes[index,0],axes[index,1])
+        if index in region_ld_legends:
+            region_ld_legend = True
+        else:
+            region_ld_legend = False
         #################################################################
         if index==0:
             # plot last m and gene track 
@@ -164,6 +174,7 @@ def plot_stacked_mqq(objects,
                             fontsize=fontsize,
                             font_family=font_family,
                             region_lead_grid=False,
+                            region_ld_legend=region_ld_legend,
                             gtf_path="default",
                             rr_ylabel=False,
                             figax=figax,
@@ -192,7 +203,7 @@ def plot_stacked_mqq(objects,
                             font_family=font_family,
                             mode=mode,
                             rr_ylabel=False,
-                            region_ld_legend=False,
+                            region_ld_legend=region_ld_legend,
                             gtf_path=None,
                             figax=figax,
                             _get_region_lead=True,
@@ -257,7 +268,7 @@ def plot_stacked_mqq(objects,
     
     ##########################################################################################################################################
     # draw the line for lead variants
-    _draw_grid_line_for_lead_variants(mode, lead_variants_is,lead_variants_is_color, n_plot, axes, region_lead_grid_line,region_chromatin_files)
+    _draw_grid_line_for_lead_variants(mode, lead_variants_is,lead_variants_is_color, n_plot, axes, region_lead_grid_line,region_chromatin_files,region_lead_grids)
     
     ##########################################################################################################################################  
     _drop_old_y_labels(axes, n_plot)
@@ -275,21 +286,24 @@ def _drop_old_y_labels(axes, n_plot):
     for index in range(n_plot):
         axes[index].set_ylabel("")
 
-def _draw_grid_line_for_lead_variants(mode, lead_variants_is,lead_variants_is_color, n_plot, axes, region_lead_grid_line,region_chromatin_files):
+def _draw_grid_line_for_lead_variants(mode, lead_variants_is,lead_variants_is_color, n_plot, axes, region_lead_grid_line,region_chromatin_files,region_lead_grids):
     if len(region_chromatin_files)>0:
         n_plot_and_track = n_plot+2
     else:
         n_plot_and_track = n_plot+1
+    
+    plotted=[None]
     if mode=="r":
         for index, sig_is in lead_variants_is.items():
-            for j, sig_i in enumerate(sig_is):
-                try:
-                    region_lead_grid_line["color"]=lead_variants_is_color[index][j]
-                except:
-                    pass
-                if sig_i is not None:
-                    for each_axis_index in range(n_plot_and_track):    
-                        axes[each_axis_index].axvline(x=sig_i, zorder=2,**region_lead_grid_line)
+            if index in region_lead_grids:
+                for j, sig_i in enumerate(sig_is):
+                    try:
+                        region_lead_grid_line["color"] = lead_variants_is_color[index][j]
+                    except:
+                        pass
+                    if sig_i not in plotted:
+                        for each_axis_index in range(n_plot_and_track):  
+                            axes[each_axis_index].axvline(x=sig_i, zorder=2,**region_lead_grid_line)
 
 def _add_new_y_label(mode, fig, gene_track_height,n_plot,subplot_height ,fontsize,font_family):
     gene_track_height_ratio = gene_track_height/(gene_track_height + n_plot*subplot_height)

{gwaslab-3.4.47 → gwaslab-3.4.48/src/gwaslab.egg-info}/PKG-INFO

@@ -1,6 +1,6 @@
 Metadata-Version: 2.1
 Name: gwaslab
-Version: 3.4.47
+Version: 3.4.48
 Summary: A collection of handy tools for GWAS SumStats
 Author-email: Yunye <yunye@gwaslab.com>
 Project-URL: Homepage, https://cloufield.github.io/gwaslab/
@@ -51,7 +51,7 @@ Warning: Known issues of GWASLab are summarized in [https://cloufield.github.io/
 ### install via pip
 
 ```
-pip install gwaslab==3.4.45
+pip install gwaslab==3.4.46
 ```
 
 ```python

{gwaslab-3.4.47 → gwaslab-3.4.48}/src/gwaslab.egg-info/SOURCES.txt

@@ -34,6 +34,7 @@ src/gwaslab/ldsc_sumstats.py
 src/gwaslab/qc_check_datatype.py
 src/gwaslab/qc_fix_sumstats.py
 src/gwaslab/run_script.py
+src/gwaslab/util_abf_finemapping.py
 src/gwaslab/util_ex_calculate_ldmatrix.py
 src/gwaslab/util_ex_calculate_prs.py
 src/gwaslab/util_ex_gwascatalog.py