Merge pull request #92 from nschcolnicov/database_fix

Added handling for custom databases

Author: lpantano

HISTORY.md

@@ -1,3 +1,7 @@
+0.4.29
+
+* Update handling of --database argument [#90](https://github.com/miRTop/mirtop/issues/90)
+
 0.4.28
 
 * fix random order in Variant field [#84](https://github.com/miRTop/mirtop/issues/83)

mirtop/command_line.py

@@ -14,6 +14,7 @@
 from mirtop.libs import spikeins
 from mirtop.gff import update
 from mirtop.sql import sql 
+from mirtop.mirna import mapper
 import mirtop.libs.logger as mylog
 
 import time
@@ -25,6 +26,9 @@ def main(**kwargs):
                       kwargs['args'].print_debug)
     logger = mylog.getLogger(__name__)
     start = time.time()
+    if not hasattr(kwargs["args"], "database"):
+        if ("sql" not in kwargs and "stats" not in kwargs and "update" not in kwargs and "validate" not in kwargs):
+            kwargs["args"].database = mapper.guess_database(kwargs["args"])
 
     if "gff" in kwargs:
         logger.info("Run annotation")

mirtop/exporter/isomirs.py

@@ -38,7 +38,7 @@ def convert(args):
 def _convert_file(gff, args):
     sep = "\t"
     precursors = fasta.read_precursor(args.hairpin, args.sps)
-    matures = mapper.read_gtf_to_precursor(args.gtf)
+    matures = mapper.read_gtf_to_precursor(args.gtf, args.database)
     variant_header = sep.join(['mism', 'add', 't5', 't3'])
 
     gff_file = open(gff, 'r')

mirtop/exporter/vcf.py

@@ -25,7 +25,7 @@ def convert(args):
     for fn in args.files:
         out_file = op.join(args.out, "%s.vcf" % op.splitext(op.basename(fn))[0])
         logger.info("Reading %s" % fn)
-        create_vcf(fn, args.hairpin, args.gtf, out_file)
+        create_vcf(fn, args.hairpin, args.gtf, out_file, args.database)
         logger.info("VCF generated %s" % out_file)
 
 
@@ -121,7 +121,7 @@ def cigar_2_key(cigar, readseq, refseq, pos, var5p, var3p, parent_ini_pos, paren
     return(key_pos, key_var, ref, alt)
 
 
-def create_vcf(mirgff3, precursor, gtf, vcffile):
+def create_vcf(mirgff3, precursor, gtf, vcffile, database):
     """
     Args:
         'mirgff3(str)': File with mirGFF3 format that will be converted
@@ -178,7 +178,7 @@ def create_vcf(mirgff3, precursor, gtf, vcffile):
     n_noSNP = 0
     no_var = 0
     hairpins = read_precursor(precursor)
-    gff3 = read_gtf_to_precursor(gtf)
+    gff3 = read_gtf_to_precursor(gtf, database)
     gtf_dic = read_gtf_to_mirna(gtf)
     for line in range(0, len(gff3_data)):
         if not gff3_data[line]:

mirtop/gff/__init__.py

@@ -21,11 +21,14 @@ def reader(args):
         read.reader(args)
         return None
     samples = []
-    database = mapper.guess_database(args)
+    if args.database is None:
+        database = mapper.guess_database(args)
+    else:
+        database = args.database
     args.database = database
     precursors = fasta.read_precursor(args.hairpin, args.sps)
     args.precursors = precursors
-    matures = mapper.read_gtf_to_precursor(args.gtf)
+    matures = mapper.read_gtf_to_precursor(args.gtf,database)
     args.matures = matures
     # TODO check numbers of miRNA and precursors read
     # TODO print message if numbers mismatch

mirtop/gff/convert.py

@@ -30,7 +30,7 @@ def convert_gff_counts(args):
                       'iso_add3p', 'iso_snp']
     if args.add_extra:
         precursors = fasta.read_precursor(args.hairpin, args.sps)
-        matures = mapper.read_gtf_to_precursor(args.gtf)
+        matures = mapper.read_gtf_to_precursor(args.gtf, args.database)
         variant_header = variant_header + ['iso_5p_nt', 'iso_3p_nt', 'iso_add3p_nt', 'iso_snp_nt']
 
     logger.info("INFO Reading GFF file %s", args.gff)

mirtop/gff/read.py

@@ -20,7 +20,7 @@ def reader(args):
     args.database = database
     precursors = fasta.read_precursor(args.hairpin, args.sps)
     args.precursors = precursors
-    matures = mapper.read_gtf_to_precursor(args.gtf)
+    matures = mapper.read_gtf_to_precursor(args.gtf, args.database)
     args.matures = matures
     # TODO check numbers of miRNA and precursors read
     # TODO print message if numbers mismatch

mirtop/importer/prost.py

@@ -41,7 +41,7 @@ def read_file(fn, hairpins, database, mirna_gtf):
     reads = defaultdict(hits)
     sample = os.path.splitext(os.path.basename(fn))[0]
     genomics = mapper.read_gtf_to_mirna(mirna_gtf)
-    matures = mapper.read_gtf_to_precursor(mirna_gtf)
+    matures = mapper.read_gtf_to_precursor(mirna_gtf, database)
     non_mirna = 0
     non_chromosome_mirna = 0
     outside_mirna = 0

mirtop/mirna/mapper.py

@@ -20,6 +20,8 @@ def guess_database(args):
 
     TODO: this needs to be generic to other databases.
     """
+    if not hasattr(args, "database"):
+        args.database = None
     return _guess_database_file(args.gtf, args.database)
 
 
@@ -143,7 +145,7 @@ def read_gtf_chr2mirna2(gtf):  # to remove
     return db_mir
 
 
-def read_gtf_to_precursor(gtf):
+def read_gtf_to_precursor(gtf,database):
     """
     Load GTF file with precursor positions on genome
     Return dict with key being precursor name and
@@ -161,15 +163,26 @@ def read_gtf_to_precursor(gtf):
     """
     if not gtf:
         return gtf
-    if _guess_database_file(gtf).find("miRBase") > -1:
+    if _guess_database_file(gtf,database).find("miRBase") > -1:
         mapped = read_gtf_to_precursor_mirbase(gtf)
-    elif _guess_database_file(gtf).find("MirGeneDB") > -1:
+    elif _guess_database_file(gtf,database).find("MirGeneDB") > -1:
         mapped = read_gtf_to_precursor_mirgenedb(gtf)
     else:
         logger.info("Database different than miRBase or MirGeneDB")
         logger.info("If you get an error when loading,")
         logger.info("report it to https://github.com/miRTop/mirtop/issues")
-        mapped = read_gtf_to_precursor_mirbase(gtf)
+        try:
+            mapped = read_gtf_to_precursor_mirbase(gtf)
+            return mapped
+        except Exception as e:
+            print(f"Failed to parse with Mirbase: {e}")
+            try:
+                mapped = read_gtf_to_precursor_mirgenedb(gtf)
+                return mapped
+            except Exception as e:
+                print(f"Failed to parse with Mirgenedb: {e}")
+                raise ValueError(f"There is no parser available for the database that you used: {database}")
+
     return mapped
 
 

requirements.txt

@@ -3,5 +3,5 @@ pybedtools
 pandas
 biopython
 pyyaml
-pybedtools
 six
+pytest