gwaslab 3.4.45__py3-none-any.whl → 3.4.47__py3-none-any.whl

gwaslab/__init__.py

@@ -44,4 +44,5 @@ from gwaslab.viz_plot_trumpetplot import plot_power
 from gwaslab.viz_plot_trumpetplot import plot_power_x
 from gwaslab.util_ex_process_h5 import process_vcf_to_hfd5
 from gwaslab.util_ex_run_susie import _run_susie_rss as run_susie_rss
-from gwaslab.io_read_tabular import _read_tabular as read_tabular
+from gwaslab.io_read_tabular import _read_tabular as read_tabular
+from gwaslab.util_in_meta import meta_analyze

gwaslab/bd_common_data.py

@@ -298,6 +298,28 @@ def gtf_to_protein_coding(gtfpath,log=Log(),verbose=True):
 
     return protein_coding_path
 
+def gtf_to_all_gene(gtfpath,log=Log(),verbose=True):
+    all_gene_path = gtfpath[:-6]+"all_genes.gtf.gz"
+    # if not existing, extract protein coding records and output to a new file
+    if not path.isfile(all_gene_path):
+        
+        # get gene list
+        log.write(" - Extracting genes from {}".format(gtfpath),verbose=verbose)
+        gtf = read_gtf(gtfpath,usecols=["feature","gene_biotype","gene_id","gene_name"])
+        gene_list = gtf.loc[gtf["feature"]=="gene","gene_id"].values
+        log.write(" - Loaded {} genes.".format(len(gene_list)),verbose=verbose)
+        
+        # extract entry using csv
+        gtf_raw = pd.read_csv(gtfpath,sep="\t",header=None,comment="#",dtype="string")
+        gtf_raw["_gene_id"] = gtf_raw[8].str.extract(r'gene_id "([\w\.-]+)"')
+        gtf_raw = gtf_raw.loc[ gtf_raw["_gene_id"].isin(gene_list) ,:]
+        gtf_raw = gtf_raw.drop("_gene_id",axis=1)
+        
+        log.write(" - Extracted records are saved to : {} ".format(all_gene_path),verbose=verbose)
+        gtf_raw.to_csv(all_gene_path, header=None, index=None, sep="\t")
+
+    return all_gene_path
+
 ####################################################################################################################   
 # From BioPython: https://github.com/biopython/biopython/blob/c5a6b1374267d769b19c1022b4b45472316e78b4/Bio/Seq.py#L36
 def _maketrans(complement_mapping):

gwaslab/g_Sumstats.py

@@ -121,6 +121,7 @@ class Sumstats():
              snpr2=None,
              status=None,
              other=[],
+             usekeys=None,
              direction=None,
              verbose=True,
              study="Study_1",
@@ -200,6 +201,7 @@ class Sumstats():
           trait=trait,
           status=status,
           other=other,
+          usekeys=usekeys,
           verbose=verbose,
           readargs=readargs,
           log=self.log)

gwaslab/g_version.py

@@ -15,16 +15,16 @@ def _get_version():
 def gwaslab_info():
     # version meta information
     dic={
-   "version":"3.4.45",
-   "release_date":"20240509"
+   "version":"3.4.47",
+   "release_date":"20240703"
     }
     return dic   
 
-def _checking_plink_version(v=2,log=Log(), verbose=True):
-    if v==1:
-        which_plink_script = "plink --version"
-    elif v==2:
-        which_plink_script = "plink2 --version" 
+def _checking_plink_version(plink=None,plink2=None,log=Log(), verbose=True):
+    if plink is not None:
+        which_plink_script = "{} --version".format(plink)
+    elif plink2 is not None:
+        which_plink_script = "{}  --version".format(plink2)
     output = subprocess.check_output(which_plink_script, stderr=subprocess.STDOUT, shell=True,text=True)
     log.write(" -PLINK version: {}".format(output.strip()))
     return log

gwaslab/hm_harmonize_sumstats.py

@@ -868,8 +868,9 @@ def parallelizeassignrsid(sumstats, path, ref_mode="vcf",snpid="SNPID",rsid="rsI
         if is_enough_info == False: return sumstats
         ############################################################################################
         
-        standardized_normalized = sumstats["STATUS"].str.match("\w\w\w[0][01234]\w\w", case=False, flags=0, na=False)
-        
+        #standardized_normalized = sumstats["STATUS"].str.match("\w\w\w[0][01234]\w\w", case=False, flags=0, na=False)
+        standardized_normalized = sumstats["STATUS"] == sumstats["STATUS"]
+
         if rsid not in sumstats.columns:
             sumstats[rsid]=pd.Series(dtype="string")
             

gwaslab/io_preformat_input.py

@@ -55,6 +55,7 @@ def preformat(sumstats,
           trait=None,
           build=None,
           other=[],
+          usekeys=None,
           verbose=False,
           readargs=None,
           log=None):
@@ -65,6 +66,11 @@ def preformat(sumstats,
     dtype_dictionary ={}    
     
  #######################################################################################################################################################
+    # workflow: 
+    # 1. formatbook
+    # 2. user specified header
+    # 3. usekeys
+    
     if fmt is not None:
         # loading format parameters
         log.write("Start to load format from formatbook....",verbose=verbose)
@@ -129,6 +135,8 @@ def preformat(sumstats,
         
         ################################################
         for key,value in rename_dictionary.items():
+            # check avaiable keys  key->raw header
+            # usecols : a list of raw headers to load from file/DataFrame 
             if key in raw_cols:
                 usecols.append(key)
             if value in ["EA","NEA"]:
@@ -137,7 +145,7 @@ def preformat(sumstats,
                 dtype_dictionary[value]="string"     
     
     except ValueError:
-        raise ValueError("Please input a path or a pd.DataFrame, and make sure the columns you specified are in the file.")
+        raise ValueError("Please input a path or a pd.DataFrame, and make sure the separator is correct and the columns you specified are in the file.")
         
     ###################################################################################################################################################
     ## check columns/datatype to use 
@@ -276,6 +284,19 @@ def preformat(sumstats,
         else:
             study = raw_cols[9]
             usecols =  usecols + [study]
+
+    if usekeys is not None:
+    # extract only specified keys
+        usecols_new =[]
+        for i in usekeys:
+            for k, v in rename_dictionary.items():
+                if i == v:
+                    usecols_new.append(k)
+        usecols_valid =[]
+        for i in usecols_new:
+            if i in usecols:
+                usecols_valid.append(i)
+        usecols = usecols_valid
  #loading data ##########################################################################################################
     
     try:

gwaslab/io_to_formats.py

@@ -212,8 +212,10 @@ def tofmt(sumstats,
     log.write(" -Start outputting sumstats in "+fmt+" format...")
     
     if "CHR" in sumstats.columns:
+        # output X,Y,MT instead of 23,24,25
         if xymt_number is False and pd.api.types.is_integer_dtype(sumstats["CHR"]):
             sumstats["CHR"]= sumstats["CHR"].map(get_number_to_chr(xymt=xymt,prefix=chr_prefix))
+        # add prefix to CHR
         elif chr_prefix is not None:
             sumstats["CHR"]= chr_prefix + sumstats["CHR"].astype("string")
 
@@ -437,17 +439,20 @@ def _configure_output_cols_and_args(sumstats, rename_dictionary, cols, no_status
             ouput_cols.append(i)  
     
     # + additional cols and remove duplicated
-    ouput_cols = list(set(ouput_cols + cols)) 
+    ouput_cols_final = []
+    for i in ouput_cols + cols:
+        if i not in ouput_cols_final:
+            ouput_cols_final.append(i)
     
     # remove STATUS
     try:
         if no_status == True:
-            ouput_cols.remove("STATUS")
+            ouput_cols_final.remove("STATUS")
     except:
         pass
     
     #filter and rename to target fromat headers
-    sumstats = sumstats[ouput_cols]
+    sumstats = sumstats[ouput_cols_final]
     sumstats = sumstats.rename(columns=rename_dictionary) 
 
     # configure target format args and reorder columns

gwaslab/qc_fix_sumstats.py

@@ -1061,6 +1061,13 @@ def check_range(sumstats, var_range, header, coltocheck, cols_to_check, log, ver
             if sum(is_low_p) >0:
                 log.warning("Extremely low P detected (P=0 or P < minimum positive value of float64) : {}".format(sum(is_low_p)))
                 log.warning("Please consider using MLOG10P instead.")
+        
+        if header=="INFO":
+            is_high_info =  sumstats["INFO"]>1 
+            if sum(is_high_info) >0:
+                log.warning("High INFO detected (INFO>1) : {}".format(sum(is_high_info)))
+                log.warning("max(INFO): {}".format(sumstats["INFO"].max()))
+                log.warning("Please check if this is as expected.")
 
         if sum(~is_valid)>0:
             try:
@@ -1102,7 +1109,7 @@ def sanitycheckstats(sumstats,
                      HR=(-100,100),
                      HR_95L=(0,float("Inf")),
                      HR_95U=(0,float("Inf")),
-                     info=(0,1),
+                     info=(0,2),
                      float_tolerence = 1e-7,
                      verbose=True,
                      log=Log()):

gwaslab/util_ex_calculate_ldmatrix.py

@@ -17,6 +17,8 @@ def tofinemapping(sumstats,
                   vcf=None, 
                   loci=None,
                   out="./",
+                  plink="plink",
+                  plink2="plink2",
                   windowsizekb=1000,
                   n_cores=1, 
                   mode="r",
@@ -56,6 +58,9 @@ def tofinemapping(sumstats,
     else:
         sig_df = sumstats.loc[sumstats["SNPID"].isin(loci),:]
     
+    log.write(" -plink1.9 path: {}".format(plink),verbose=verbose)
+    log.write(" -plink2 path: {}".format(plink2),verbose=verbose)
+
     # Drop duplicate!!!!
     log.write(" -Dropping duplicated SNPIDs...",verbose=verbose)
     sumstats = sumstats.drop_duplicates(subset=["SNPID"]).copy()
@@ -68,11 +73,13 @@ def tofinemapping(sumstats,
     if exclude_hla==True:
         sig_df = _exclude_hla(sig_df, log=log, verbose=verbose)
     
+    sig_df = sig_df.reset_index()
+    
     ## for each lead variant 
     for index, row in sig_df.iterrows():
         # extract snplist in each locus
         gc.collect()
-
+        log.write(" -Locus #{}---------------------------------------------------------------".format(index+1))
         log.write(" -Processing locus with lead variant {} at CHR {} POS {} ...".format(row["SNPID"],row["CHR"],row["POS"]))
         locus_sumstats = _extract_variants_in_locus(sumstats, windowsizekb, locus = (row["CHR"],row["POS"]))
         
@@ -84,7 +91,10 @@ def tofinemapping(sumstats,
                                                                     n_cores=n_cores, 
                                                                     log=log,
                                                                     load_bim=True,
-                                                                    overwrite=overwrite,**kwargs)
+                                                                    overwrite=overwrite,
+                                                                    plink=plink,
+                                                                    plink2=plink2,
+                                                                    **kwargs)
 
         ## check available snps with reference file
         matched_sumstats = _align_sumstats_with_bim(row=row, 
@@ -114,7 +124,10 @@ def tofinemapping(sumstats,
                                                             windowsizekb=windowsizekb,
                                                             out=out,
                                                             plink_log=plink_log,
-                                                            log=log,filetype=filetype,
+                                                            log=log,
+                                                            filetype=filetype,
+                                                            plink=plink,
+                                                            plink2=plink2,
                                                             verbose=verbose)
     
     
@@ -143,12 +156,12 @@ def tofinemapping(sumstats,
 
 
 
-def _calculate_ld_r(study, matched_sumstats_snpid, row, bfile_prefix, n_cores, windowsizekb,out,plink_log,log,memory,mode,filetype,verbose=True):
+def _calculate_ld_r(study, matched_sumstats_snpid, row, bfile_prefix, n_cores, windowsizekb,out,plink_log,log,memory,mode,filetype,plink,plink2,verbose=True):
     '''
     Calculate LD r matrix by calling PLINK; return file name and log
     '''
     log.write(" -Start to calculate LD r matrix...",verbose=verbose)
-    log = _checking_plink_version(v=1, log=log)
+    log = _checking_plink_version(plink=plink, log=log)
     if "@" in bfile_prefix:
         bfile_to_use = bfile_prefix.replace("@",str(row["CHR"]))
     else:
@@ -165,7 +178,7 @@ def _calculate_ld_r(study, matched_sumstats_snpid, row, bfile_prefix, n_cores, w
             raise ValueError("Please use bfile instead of pfile for PLINK1.")
         
         script_vcf_to_bfile = """
-        plink \
+        {} \
             --bfile {} \
             --keep-allele-order \
             --extract {} \
@@ -175,7 +188,7 @@ def _calculate_ld_r(study, matched_sumstats_snpid, row, bfile_prefix, n_cores, w
             --threads {} {}\
             --write-snplist \
             --out {}
-        """.format(bfile_to_use, snplist_path , row["CHR"], mode, n_cores, memory_flag if memory is not None else "", output_prefix)
+        """.format(plink, bfile_to_use, snplist_path , row["CHR"], mode, n_cores, memory_flag if memory is not None else "", output_prefix)
 
         try:
             output = subprocess.check_output(script_vcf_to_bfile, stderr=subprocess.STDOUT, shell=True,text=True)

gwaslab/util_ex_calculate_prs.py

@@ -18,6 +18,8 @@ def _calculate_prs(sumstats,
           memory=None, 
           overwrite=False,
           mode=None,delete=True,
+          plink="plink",
+          plink2="plink2",
           log=Log(),**kwargs):
     
     #matching_alleles
@@ -30,14 +32,18 @@ def _calculate_prs(sumstats,
         chrlist.sort()
         plink_log = ""
         #process reference fileWWW
-        bfile_prefix, plink_log, ref_bim, filetype = _process_plink_input_files(  chrlist=chrlist,
+        bfile_prefix, plink_log, ref_bim, filetype = _process_plink_input_files(  
+                                                                    chrlist=chrlist,
                                                                     bfile=bfile, 
                                                                     vcf=vcf, 
                                                                     plink_log=plink_log,
                                                                     n_cores=n_cores, 
                                                                     log=log,
                                                                     load_bim=False,
-                                                                    overwrite=overwrite,**kwargs)
+                                                                    overwrite=overwrite,
+                                                                    plink=plink,
+                                                                    plink2=plink2,
+                                                                    **kwargs)
         score_file_path_list =[]
         for index, chrom in enumerate(chrlist): 
             chr_sumstats = sumstats.loc[sumstats["CHR"]==chrom,:].copy()
@@ -61,7 +67,7 @@ def _calculate_prs(sumstats,
                                plink_log=plink_log, 
                                log=log, 
                                memory=memory, 
-                               mode=mode,filetype=filetype)
+                               mode=mode,filetype=filetype,plink2=plink2)
             score_file_path_list.append(score_file_path)
             if delete == True:
                 os.remove(model_path)
@@ -71,10 +77,10 @@ def _calculate_prs(sumstats,
 
 
 
-def _run_calculate_prs(study, chrom , model_path, bfile_prefix, n_cores, out, plink_log, log, memory,filetype, mode=None):
+def _run_calculate_prs(study, chrom , model_path, bfile_prefix, n_cores, out, plink_log, log, memory,filetype, plink2,mode=None):
     
     log.write(" -Start to calculate PRS for Chr {}...".format(chrom))
-    _checking_plink_version(v=2, log=log)
+    _checking_plink_version(plink2=plink2, log=log)
     
     if "@" in bfile_prefix:
         bpfile_to_use = bfile_prefix.replace("@",str(chrom))
@@ -92,13 +98,13 @@ def _run_calculate_prs(study, chrom , model_path, bfile_prefix, n_cores, out, pl
         memory_flag = "--memory {}".format(memory)
     
     script_vcf_to_bfile = """
-    plink2 \
+    {} \
         {} \
         --score {} 1 2 3 header {} cols=+scoresums,+denom ignore-dup-ids \
         --chr {} \
         --threads {} {}\
         --out {}
-    """.format(file_flag, model_path ,  mode if mode is not None else "", chrom, n_cores, memory_flag if memory is not None else "", output_prefix)
+    """.format(plink2, file_flag, model_path ,  mode if mode is not None else "", chrom, n_cores, memory_flag if memory is not None else "", output_prefix)
 
     try:
         output = subprocess.check_output(script_vcf_to_bfile, stderr=subprocess.STDOUT, shell=True,text=True)

gwaslab/util_ex_process_ref.py

@@ -20,7 +20,9 @@ def _process_plink_input_files(chrlist,
                                bgen_mode="ref-first", 
                                convert="bfile", 
                                memory=None, 
-                               load_bim=False):
+                               load_bim=False,
+                               plink="plink",
+                               plink2="plink2"):
     """
     Process input files (bfile,pfile,vcf,bgen) to either PLINK1 bed/bim/fam or PLINK2 pgen/psam/pvar. 
     
@@ -66,7 +68,9 @@ def _process_plink_input_files(chrlist,
                                                             convert=convert, 
                                                             memory=memory, 
                                                             overwrite=overwrite, 
-                                                            load_bim=load_bim)
+                                                            load_bim=load_bim,
+                                                            plink=plink,
+                                                            plink2=plink2)
         filetype = convert
     elif filetype == "bgen": 
         ref_file_prefix, plink_log, ref_bims = _process_bgen(ref_file_prefix=ref_file_prefix, 
@@ -81,7 +85,9 @@ def _process_plink_input_files(chrlist,
                                                             convert=convert, 
                                                             memory=memory, 
                                                             overwrite=overwrite, 
-                                                            load_bim=load_bim)
+                                                            load_bim=load_bim,
+                                                            plink=plink,
+                                                            plink2=plink2)
         filetype = convert
     return ref_file_prefix, plink_log, ref_bims, filetype
 
@@ -199,11 +205,13 @@ def _process_vcf(ref_file_prefix,
                  convert="bfile", 
                  memory=None, 
                  overwrite=False, 
-                 load_bim=False):
+                 load_bim=False,
+                 plink="plink",
+                 plink2="plink2"):
     log.write(" -Processing VCF : {}...".format(ref_file_prefix))
     
     #check plink version
-    log = _checking_plink_version(v=2,log=log)
+    log = _checking_plink_version(plink2=plink2,log=log)
     
     # file path prefix to return
     if is_wild_card==True:
@@ -243,14 +251,15 @@ def _process_vcf(ref_file_prefix,
         #if not existing or overwrite is True
         if (not is_file_exist) or overwrite:
             script_vcf_to_bfile = """
-            plink2 \
+            {} \
                 --vcf {} \
                 --chr {} \
                 {} \
                 --rm-dup force-first \
                 --threads {}{}\
                 --out {}
-            """.format(vcf_to_load, 
+            """.format(plink2,
+                        vcf_to_load, 
                        i, 
                        make_flag,
                        n_cores, memory_flag,
@@ -288,11 +297,13 @@ def _process_bgen(ref_file_prefix,
                   convert="bfile", 
                   memory=None, 
                   overwrite=False, 
-                  load_bim=False):
+                  load_bim=False,
+                  plink="plink",
+                 plink2="plink2"):
     log.write(" -Processing BGEN files : {}...".format(ref_file_prefix))
     
     #check plink version
-    log = _checking_plink_version(v=2,log=log)
+    log = _checking_plink_version(log=log,plink2=plink2)
     
     # file path prefix to return
     if is_wild_card==True:
@@ -338,14 +349,14 @@ def _process_bgen(ref_file_prefix,
         #if not existing or overwrite is True
         if (not is_file_exist) or overwrite:
             script_vcf_to_bfile = """
-            plink2 \
+            {} \
                 --bgen {} {} {}\
                 --chr {} \
                 {} \
                 --rm-dup force-first \
                 --threads {}{}\
                 --out {}
-            """.format(bgen_to_load, bgen_mode, sample_flag,
+            """.format(plink2,bgen_to_load, bgen_mode, sample_flag,
                        i, 
                        make_flag,
                        n_cores, memory_flag,

gwaslab/util_in_filter_value.py

@@ -10,6 +10,7 @@ from gwaslab.g_vchange_status import vchange_status
 from gwaslab.qc_fix_sumstats import sortcoordinate
 from gwaslab.qc_fix_sumstats import start_to
 from gwaslab.qc_fix_sumstats import finished
+from gwaslab.qc_fix_sumstats import _process_build
 from gwaslab.hm_harmonize_sumstats import is_palindromic
 
 import gc
@@ -430,8 +431,43 @@ def _filter_snp(sumstats, mode="in", ea="EA",nea="NEA", log=Log(),verbose=True):
     log.write("Finished filtering SNPs.",verbose=verbose)
     return snp
 
-def _exclude_hla(sumstats, chrom="CHR", pos="POS", lower=25000000 ,upper=34000000 ,log=Log(), verbose=True):
-    
+def _exclude_hla(sumstats, chrom="CHR", pos="POS", lower=None ,upper=None, build=None, mode="xmhc", log=Log(), verbose=True):
+    
+    if build is not None:
+        build = _process_build(build = build,log = log,verbose = verbose)
+        # xMHC : HIST1H2AA ~ 7.6mb ~ RPL12P1
+        # reference: Horton, R., Wilming, L., Rand, V., Lovering, R. C., Bruford, E. A., Khodiyar, V. K., ... & Beck, S. (2004). Gene map of the extended human MHC. Nature Reviews Genetics, 5(12), 889-899.
+        # hg38:  25,726,063 ~ 33,400,644
+        # hg19 : 25,726,291 ~ 33,368,421
+
+        # HLA : GABBR1 ~ 3.78mb ~ KIFC1
+        # reference: Shiina, T., Hosomichi, K., Inoko, H., & Kulski, J. K. (2009). The HLA genomic loci map: expression, interaction, diversity and disease. Journal of human genetics, 54(1), 15-39. 
+        # hg38:  29,602,238 ~ 33,409,896
+        # hg19:  29,570,015 ~ 33,377,673
+
+        if build == "19":
+            if mode =="xmhc":
+                lower=25000000
+                upper=34000000
+            if mode =="hla" or mode =="mhc":
+                lower=29500000
+                upper=33500000
+        if build == "38":
+            if mode =="xmhc":
+                lower=25000000
+                upper=34000000
+            if mode =="hla" or mode =="mhc":
+                lower=29500000
+                upper=33500000
+    else:
+        # -> 25,000,000 ~ 34,000,000
+        if mode =="xmhc":
+            lower=25000000
+            upper=34000000
+        if mode =="hla" or mode =="mhc":
+            lower=29500000
+            upper=33500000
+        
     raw_len = len(sumstats)
     
     if str(sumstats[chrom].dtype) == "string":

gwaslab/util_in_get_sig.py

@@ -11,6 +11,7 @@ from gwaslab.bd_common_data import get_chr_to_number
 from gwaslab.bd_common_data import get_number_to_chr
 from gwaslab.bd_common_data import get_chr_to_NC
 from gwaslab.bd_common_data import gtf_to_protein_coding
+from gwaslab.bd_common_data import gtf_to_all_gene
 from gwaslab.bd_download import check_and_download
 from gwaslab.util_ex_gwascatalog import gwascatalog_trait
 from gwaslab.qc_fix_sumstats import check_dataframe_shape
@@ -38,6 +39,7 @@ def getsig(insumstats,
            wc_correction=False,
            build="19",
            source="ensembl",
+           gtf_path=None,
            verbose=True):
     """
     Extract the lead variants using a sliding window. P or MLOG10P will be used and converted to SCALEDP for sorting. 
@@ -172,6 +174,7 @@ def getsig(insumstats,
                xymt=xymt,
                build=build,
                source=source,
+               gtf_path=gtf_path,
                verbose=verbose)
         
     # drop internal id
@@ -253,6 +256,7 @@ def annogene(
            xymt=["X","Y","MT"],
            build="19",
            source="ensembl",
+           gtf_path=None,
            verbose=True):
     
     log.write("Start to annotate variants with nearest gene name(s)...", verbose=verbose)
@@ -267,8 +271,13 @@ def annogene(
             #| gzip >Homo_sapiens.GRCh37.75.processed.chr.gtf.gz     
             
             #gtf_path = check_and_download("ensembl_hg19_gtf_protein_coding")
-            gtf_path = check_and_download("ensembl_hg19_gtf")
-            gtf_path = gtf_to_protein_coding(gtf_path,log=log,verbose=verbose)
+            if gtf_path is None:
+                gtf_path = check_and_download("ensembl_hg19_gtf")
+                gtf_path = gtf_to_protein_coding(gtf_path,log=log,verbose=verbose)
+            else:
+                log.write(" -Using user-provided gtf:{}".format(gtf_path))
+                gtf_path = gtf_to_all_gene(gtf_path,log=log,verbose=verbose)
+
             gtf_db_path = gtf_path[:-2]+"db"
             
             data = Genome(
@@ -283,8 +292,13 @@ def annogene(
         elif build=="38":
             log.write(" -Assigning Gene name using ensembl_hg38_gtf for protein coding genes", verbose=verbose)
             #gtf_path = check_and_download("ensembl_hg38_gtf_protein_coding")
-            gtf_path = check_and_download("ensembl_hg38_gtf")
-            gtf_path = gtf_to_protein_coding(gtf_path,log=log,verbose=verbose)
+            if gtf_path is None:
+                gtf_path = check_and_download("ensembl_hg38_gtf")
+                gtf_path = gtf_to_protein_coding(gtf_path,log=log,verbose=verbose)
+            else:
+                log.write(" -Using user-provided gtf:{}".format(gtf_path))
+                gtf_path = gtf_to_all_gene(gtf_path,log=log,verbose=verbose)
+            
             gtf_db_path = gtf_path[:-2]+"db"
             data = Genome(
                 reference_name='GRCh38',
@@ -300,8 +314,13 @@ def annogene(
         if build=="19":
             log.write(" -Assigning Gene name using NCBI refseq latest GRCh37 for protein coding genes", verbose=verbose)
             #gtf_path = check_and_download("refseq_hg19_gtf_protein_coding")
-            gtf_path = check_and_download("refseq_hg19_gtf")
-            gtf_path = gtf_to_protein_coding(gtf_path,log=log,verbose=verbose)
+            if gtf_path is None:
+                gtf_path = check_and_download("refseq_hg19_gtf")
+                gtf_path = gtf_to_protein_coding(gtf_path,log=log,verbose=verbose)
+            else:
+                log.write(" -Using user-provided gtf:{}".format(gtf_path))
+                gtf_path = gtf_to_all_gene(gtf_path,log=log,verbose=verbose)
+            
             gtf_db_path = gtf_path[:-2]+"db"
             data = Genome(
                 reference_name='GRCh37',
@@ -315,8 +334,13 @@ def annogene(
         elif build=="38":
             log.write(" -Assigning Gene name using NCBI refseq latest GRCh38 for protein coding genes", verbose=verbose)
             #gtf_path = check_and_download("refseq_hg38_gtf_protein_coding")
-            gtf_path = check_and_download("refseq_hg38_gtf")
-            gtf_path = gtf_to_protein_coding(gtf_path,log=log,verbose=verbose)
+            if gtf_path is None:
+                gtf_path = check_and_download("refseq_hg38_gtf")
+                gtf_path = gtf_to_protein_coding(gtf_path,log=log,verbose=verbose)
+            else:
+                log.write(" -Using user-provided gtf:{}".format(gtf_path))
+                gtf_path = gtf_to_all_gene(gtf_path,log=log,verbose=verbose)
+            
             gtf_db_path = gtf_path[:-2]+"db"
             data = Genome(
                 reference_name='GRCh38',