Merge pull request #19 from andrewjpage/direction

Validate databases

VERSION

@@ -1 +1 @@
-2.2.0
+2.2.1

socru/GATProfile.py

@@ -128,5 +128,15 @@ def orientation_binary(self):
                 total += 1 << bits_to_shift
         return total 
 
+    def orientation_array(self):
+        orientation_bools = []
+        for f in self.fragments:
+            m = re.match(r"([\d]+)'", f)
+            if m:
+                orientation_bools.append(False)
+            else:
+                orientation_bools.append(True)
+        return orientation_bools
+
 
 

socru/SocruLookup.py

@@ -3,22 +3,128 @@
 from socru.GATProfile  import GATProfile
 from socru.Profiles import Profiles
 from socru.TypeGenerator import TypeGenerator
+from socru.ValidateFragments import ValidateFragments
+from socru.Fragment import Fragment
 
 class SocruLookup:
     def __init__(self,options):
         self.fragments = options.fragments
         self.db_dir = options.db_dir
         self.verbose = options.verbose
 
-    def calc_type(self):  
-        profile_db = Profiles(os.path.join(self.db_dir, 'profile.txt'), self.verbose)
+        self.tg = self.type_generator()
+
+    def create_fragments(self, input_profile, standard_fragments):
+        frags = []
+        # get the basic details
+
+        reversed_frags = input_profile.orientation_array()
+        for i, frag_number in enumerate(input_profile.orientationless_fragments()):
+            f = Fragment([], number = frag_number, operon_forward_start = False, dna_A = False, dif = False, reversed_frag = not reversed_frags[i])
+            if str(frag_number) == str(input_profile.dnaA_fragment_number):
+                f.dna_A = True
+            if str(frag_number) == str(input_profile.dif_fragment_number):
+                f.dif = True
+            frags.append(f)
+
+        # loop over the standard and set the direction (if inverted, then invert)
+        for f in frags:
+            for s in standard_fragments:
+                if f.number == s.number:
+                    f.operon_forward_start = s.operon_forward_start
+
+            #if f.dna_A:
+            #    f.operon_forward_start = True
+            #if f.dif:
+            #    f.operon_forward_start = False
+
+            if f.reversed_frag:
+                f.operon_forward_start = not f.operon_forward_start
+
+        return frags
+
+    def find_dnaa(self, frags):
+        for i, frag in enumerate(frags):
+            if frag.dna_A:
+                return i
+
+        print("DnaA couldnt be found")
+        return -1
+
+    def find_dif(self, frags):
+        for i, frag in enumerate(frags):
+            if frag.dif:
+                return i
+
+        print("Dif couldnt be found")
+        return -1
+
+    def reorientate_ori(self, frags):
+        dnaa_index = self.find_dnaa(frags)
+        if dnaa_index == -1:
+            return frags
+
+        reorientated_frags = frags[dnaa_index: len(frags)] + frags[0: dnaa_index]
+        return reorientated_frags
 
+    # Take reference case 1 2 3 4 5 6 7
+    # assign terminus and origin
+    # assign direction to each number
+    # dnaa points outwards
+    # dif points inwards
+    # This is the standard
+    def create_standard_fragments(self, profile_db):
+        reference_profile = profile_db.gats[0]
+
+        ref_frags = []
+        # get the basic details
+        for frag_number in reference_profile.orientationless_fragments():
+            f = Fragment([], number = frag_number, operon_forward_start = False, dna_A = False, dif = False)
+            if str(frag_number) == str(profile_db.dnaA_fragment_number):
+                f.dna_A = True
+            if str(frag_number) == str(profile_db.dif_fragment_number):
+                f.dif = True
+            ref_frags.append(f)
+
+        # reorientate to dnaA
+        dna_start_frags = self.reorientate_ori(ref_frags)
+        dnaa_index = self.find_dnaa(dna_start_frags)
+        dif_index = self.find_dif(dna_start_frags)
+        if dnaa_index == -1 or dif_index == -1:
+            return []
+
+        # forward walk (reverse is already set )
+        for i in range(dnaa_index, dif_index):
+            if i >= dif_index:
+                # we have reached the end
+                continue
+            else:
+                dna_start_frags[i].operon_forward_start = True
+
+        return dna_start_frags 
+
+    def validate_profile(self, profile_db, input_profile):
+        standard_fragments = self.create_standard_fragments(profile_db)
+        frags = self.create_fragments( input_profile, standard_fragments)
+        return ValidateFragments(frags).validate()
+
+    def type_generator(self):
+        profile_db = Profiles(os.path.join(self.db_dir, 'profile.txt'), self.verbose)
+
         split_fragments = self.fragments.split('-')
-        input_profile = GATProfile(self.verbose, fragments = split_fragments)
+        input_profile = GATProfile(self.verbose, fragments = split_fragments, dnaA_fragment_number = profile_db.dnaA_fragment_number, dif_fragment_number = profile_db.dif_fragment_number)
+        input_profile.orientate_for_dnaA()
 
-        tg = TypeGenerator(profile_db,input_profile, self.verbose, True)
-        return tg.calculate_type()
+        is_profile_valid = self.validate_profile(profile_db, input_profile)
+        tg = TypeGenerator(profile_db,input_profile, self.verbose, is_profile_valid)
+        return tg
+
+    def calc_type(self):
+        return self.tg.calculate_type()
 
+    def calc_quality(self):
+        return self.tg.quality
+
     def run(self):
-        print(self.calc_type())
+        print(str(self.db_dir ) + "\t"+str(self.calc_quality())+"\t"+str(self.tg.calculate_type()))
 

socru/TypeGenerator.py

@@ -67,8 +67,7 @@ def previously_seen(self, gat_number, orientation_binary):
 
                 if self.is_frag_valid:
                     self.quality = 'GREEN' 
-                else:
-                    self.quality = 'AMBER'
+
                 return True
             else:
                 return False

socru/data/Acinetobacter_baumannii/profile.txt

@@ -1,7 +1,3 @@
 GS	Frag_1	Frag_2	Frag_3	Frag_4	Frag_5	Frag_6
 1.0	1	2	3	4	5	6
 1.1	1'	2	3	4	5	6
-1.8	1	2	3	4'	5	6
-1.9	1'	2	3	4'	5	6
-2.24	1	2	3	5'	4'	6
-2.25	1'	2	3	5'	4'	6

socru/data/Acinetobacter_sp./profile.txt.yml

@@ -1,3 +1,4 @@
+dif_fragment: 1
 dnaa_forward_orientation: true
 dnaa_fragment: 4
 reference_genome: Acinetobacter/sp./ADP1/GCF_000046845.1_ASM4684v1_genomic.fna.gz

socru/data/Bacillus_cellulasensis/profile.txt.yml

@@ -1,3 +1,4 @@
+dif_fragment: 1
 dnaa_forward_orientation: false
 dnaa_fragment: 7
 reference_genome: Bacillus/cellulasensis/NJ-M2/GCF_001431145.1_ASM143114v1_genomic.fna.gz

socru/data/Bordetella_pertussis/profile.txt

@@ -1,4 +1,4 @@
 GS	Frag_1	Frag_2	Frag_3
 1.0	1	2	3
-1.1	1'	2	3
-1.3	1'	3	2'
+1.6	1'	2	3
+1.4	1'	3	2'

socru/data/Candidatus_Kinetoplastibacterium/profile.txt.yml

@@ -1,3 +1,4 @@
+dif_fragment: 1
 dnaa_forward_orientation: true
 dnaa_fragment: 2
 reference_genome: Candidatus/Kinetoplastibacterium/crithidii__ex_Angomonas_deanei_ATCC_30255_/GCF_000319225.1_ASM31922v1_genomic.fna.gz

socru/data/Enterobacter_asburiae/profile.txt

@@ -1,4 +1,3 @@
 GS	Frag_1	Frag_2	Frag_3	Frag_4	Frag_5	Frag_6	Frag_7	Frag_8
 1.0	1	2	3	4	5	6	7	8
 1.1	1'	2	3	4	5	6	7	8
-

socru/data/Enterobacter_cloacae/profile.txt

@@ -1,4 +1,3 @@
 GS	Frag_1	Frag_2	Frag_3	Frag_4	Frag_5	Frag_6	Frag_7	Frag_8
 1.0	1	2	3	4	5	6	7	8
 1.1	1'	2	3	4	5	6	7	8
-2.57	1'	2	3	6'	5'	4'	7	8

socru/data/Enterobacter_hormaechei/profile.txt.yml

@@ -1,3 +1,4 @@
+dif_fragment: 1
 dnaa_forward_orientation: true
 dnaa_fragment: 6
 reference_genome: Enterobacter/hormaechei/ECNIH3/GCF_000750225.1_ASM75022v1_genomic.fna.gz

socru/data/Enterobacter_roggenkampii/profile.txt.yml

@@ -1,3 +1,4 @@
+dif_fragment: 1
 dnaa_forward_orientation: true
 dnaa_fragment: 6
 reference_genome: Enterobacter/roggenkampii/35734/GCF_000807415.2_ASM80741v4_genomic.fna.gz

socru/data/Enterobacter_sp./profile.txt.yml

@@ -1,3 +1,4 @@
+dif_fragment: 1
 dnaa_forward_orientation: true
 dnaa_fragment: 6
 reference_genome: Enterobacter/sp./638/GCF_000016325.1_ASM1632v1_genomic.fna.gz

socru/data/Enterococcus_faecium/profile.txt

@@ -1,9 +1,7 @@
 GS	Frag_1	Frag_2	Frag_3	Frag_4	Frag_5	Frag_6
 1.0	1	2	3	4	5	6
 1.1	1'	2	3	4	5	6
-1.8	1	2	3	4'	5	6
 2.0	6	1	2	3	4	5
-3.24	1	5'	4'	2	3	6
 4.32	1	6'	2	3	4	5
 4.33	1'	6'	2	3	4	5
 5.35	1'	6'	3	4	5	2'

socru/data/Escherichia_coli/profile.txt

@@ -1,24 +1,16 @@
 GS	Frag_1	Frag_2	Frag_3	Frag_4	Frag_5	Frag_6	Frag_7
 1.0	1	2	3	4	5	6	7
 1.1	1'	2	3	4	5	6	7
-1.2	1	2'	3	4	5	6	7
-1.24	1	2	3	4'	5'	6	7
-1.25	1'	2	3	4'	5'	6	7
-1.64	1	2	3	4	5	6	7'
-1.9	1'	2	3	4'	5	6	7
 1.123	1'	7'	6'	5'	4'	3	2'
 2.67	1'	7'	3	4	5	6	2'
-2.90	1	7'	3	4'	5'	6	2'
 3.25	1'	4'	5'	2	3	6	7
 4.96	1	7'	6'	2	3	4	5
 4.97	1'	7'	6'	2	3	4	5
 5.99	1'	7'	6'	3	4	5	2'
 6.32	1	2	6'	3	4	5	7
-7.120	1	2	3	7'	6'	5'	4'
 8.3	1'	3	4	5	6	7	2'
 9.16	1	2	5'	3	6	4	7
 10.65	1'	7'	2	3	4	5	6
 11.59	1'	4'	6'	5'	3	7	2'
-12.56	1	2	3	6'	5'	4'	7
 13.113	1'	7'	6'	5'	2	3	4
 14.64	1	2	7'	3	4	5	6

socru/data/Escherichia_coli/profile.txt.yml

@@ -1,3 +1,4 @@
+dif_fragment: 1
 dnaa_forward_orientation: false
 dnaa_fragment: 3
 reference_genome: Escherichia/coli/K-12_substr._MG1655/GCF_000005845.2_ASM584v2_genomic.fna.gz

socru/data/Flavobacterium_sp./profile.txt.yml

@@ -1,3 +1,4 @@
+dif_fragment: 1
 dnaa_forward_orientation: true
 dnaa_fragment: 3
 reference_genome: Flavobacterium/sp./MEBiC07310/GCF_003148385.1_ASM314838v1_genomic.fna.gz

socru/data/Haemophilus_influenzae/profile.txt

@@ -1,8 +1,3 @@
 GS	Frag_1	Frag_2	Frag_3	Frag_4	Frag_5	Frag_6
 1.0	1	2	3	4	5	6
-2.32	1	2	3	6'	4	5
-2.40	1	2	3	6'	4'	5
-2.48	1	2	3	6'	4	5'
-3.3	1'	3	4	6	5	2'
-4.33	1'	6'	2	3	4	5
-5.48	1	2	6'	3	4	5'
+

socru/data/Klebsiella_pneumoniae/profile.txt

@@ -1,11 +1,9 @@
 GS	Frag_1	Frag_2	Frag_3	Frag_4	Frag_5	Frag_6	Frag_7	Frag_8
 1.0	1	2	3	4	5	6	7	8
 1.1	1'	2	3	4	5	6	7	8
-1.128	1	2	3	4	5	6	7	8'
 2.1	1'	2	3	7	4	5	6	8
 3.24	1	2	5'	4'	3	6	7	8
 4.131	1'	8'	3	4	5	6	7	2'
 5.0	1	2	3	5	4	6	7	8
 6.3	1'	3	4	5	6	7	8	2'
-7.32	1	2	3	4	5	7	8	6'
 8.35	1'	6'	3	4	5	6	7	2'

socru/data/Klebsiella_pneumoniae/profile.txt.yml

@@ -1,3 +1,4 @@
+dif_fragment: 1
 dnaa_forward_orientation: false
 dnaa_fragment: 3
 reference_genome: Klebsiella/pneumoniae/NTUH-K2044/GCF_000009885.1_ASM988v1_genomic.fna.gz

socru/data/Lactobacillus_plantarum/profile.txt

@@ -1,5 +1,3 @@
 GS	Frag_1	Frag_2	Frag_3	Frag_4	Frag_5
 1.0	1	2	3	4	5
 1.1	1'	2	3	4	5
-1.8	1	2	3	4'	5
-2.8	4'	5	1	2	3

socru/data/Legionella_sainthelensi/profile.txt.yml

@@ -1,3 +1,4 @@
+dif_fragment: 1
 dnaa_forward_orientation: true
 dnaa_fragment: 2
 reference_genome: Legionella/sainthelensi/LA01-117/GCF_002848365.1_ASM284836v1_genomic.fna.gz

socru/data/Leuconostoc_gelidum/profile.txt.yml

@@ -1,3 +1,4 @@
+dif_fragment: 1
 dnaa_forward_orientation: true
 dnaa_fragment: 1
 reference_genome: Leuconostoc/gelidum/type_strain:_LMG_18811/GCF_000196855.1_ASM19685v1_genomic.fna.gz

socru/data/Listeria_monocytogenes/profile.txt

@@ -1,7 +1,4 @@
 GS	Frag_1	Frag_2	Frag_3	Frag_4	Frag_5	Frag_6
 1.0	1	2	3	4	5	6
-1.16	1	2	3	4	5'	6
-2.56	1	2	3	6'	5'	4'
 3.33	1'	6'	2	3	4	5
-4.48	1	2	3	4	6'	5'
 5.35	1'	6'	3	4	5	2'

socru/data/Mannheimia_haemolytica/profile.txt

@@ -1,5 +1,3 @@
 GS	Frag_1	Frag_2	Frag_3	Frag_4	Frag_5	Frag_6
 1.0	1	2	3	4	5	6
 2.0	1	2	5	3	4	6
-3.33	1'	6'	2	3	5	4
-4.17	1'	5'	2	3	4	6

socru/data/Neisseria_meningitidis/profile.txt

@@ -1,6 +1,3 @@
 GS	Frag_1	Frag_2	Frag_3	Frag_4
 1.0	1	2	3	4
 1.1	1'	2	3	4
-2.3	1'	3	4	2'
-2.9	1'	2	4'	3
-3.1	4	1'	2	3

socru/data/Pasteurella_multocida/profile.txt

@@ -1,5 +1,2 @@
 GS	Frag_1	Frag_2	Frag_3	Frag_4	Frag_5	Frag_6
 1.0	1	2	3	4	5	6
-2.48	1	2	3	4	6'	5'
-3.9	1'	5	6	4'	2	3
-4.24	1	2	3	6	4'	5'

socru/data/Porphyromonas_gingivalis/profile.txt

@@ -1,4 +1,2 @@
 GS	Frag_1	Frag_2	Frag_3	Frag_4
 1.0	1	2	3	4
-1.1	1'	4	3	2
-1.9	3	2	1'	4'

socru/data/Pseudomonas_aeruginosa/profile.txt

@@ -1,7 +1,5 @@
 GS	Frag_1	Frag_2	Frag_3	Frag_4
 1.0	1	2	3	4
 1.1	1'	2	3	4
-1.9	1'	2	3	4'
 2.2	1	3	4	2'
 2.3	1'	3	4	2'
-2.11	1'	3	4'	2'

socru/data/Salmonella_enterica/profile.txt.yml

@@ -1,3 +1,4 @@
+dif_fragment: 1
 dnaa_forward_orientation: false
 dnaa_fragment: 3
 reference_genome: Salmonella/enterica/subsp._enterica_serovar_Typhimurium_str._LT2/GCF_000006945.2_ASM694v2_genomic.fna.gz

socru/data/Streptococcus_pyogenes/profile.txt

@@ -3,4 +3,3 @@ GS	Frag_1	Frag_2	Frag_3	Frag_4	Frag_5	Frag_6
 1.1	1'	2	3	4	5	6
 2.32	1	2	6'	3	4	5
 3.3	1'	3	5	4	6	2'
-4.56	1	2	3	6'	5'	4'

socru/data/Streptomyces_sp./profile.txt

@@ -1,4 +1,4 @@
 GS	Frag_1	Frag_2	Frag_3	Frag_4	Frag_5	Frag_6
 1.0	1	2	3	4	5	6
-1.16	1	2	3	4	5'	6
-1.17	1'	2	3	4	5'	6
+1.47	1	2	3	4	5'	6
+1.46	1'	2	3	4	5'	6

socru/data/Vibrio_cholerae/profile.txt

@@ -1,4 +1,4 @@
 GS	Frag_1	Frag_2	Frag_3	Frag_4	Frag_5	Frag_6	Frag_7	Frag_8
 1.0	1	2	3	4	5	6	7	8
-2.251	1'	8'	3	7'	6'	5'	4'	2'
-3.251	1'	8'	7'	6'	3	5'	4'	2'
+2.4	1'	8'	3	7'	6'	5'	4'	2'
+3.4	1'	8'	7'	6'	3	5'	4'	2'

socru/data/Vibrio_harveyi/profile.txt.yml

@@ -1,3 +1,4 @@
+dif_fragment: 1
 dnaa_forward_orientation: true
 dnaa_fragment: 9
 reference_genome: Vibrio/harveyi/ATCC_33843__392_[MAV]_/GCF_000770115.1_ASM77011v2_genomic.fna.gz

socru/data/Yersinia_pseudotuberculosis/profile.txt

@@ -1,7 +1,4 @@
 GS	Frag_1	Frag_2	Frag_3	Frag_4	Frag_5	Frag_6	Frag_7
 1.0	1	2	3	4	5	6	7
 1.1	1'	2	3	4	5	6	7
-2.56	1	2	3	6'	5'	4'	7
-3.123	1'	7'	3	6'	5'	4'	2'
-4.91	1'	7'	3	5'	4'	6	2'
-5.83	1'	7'	5'	3	4	6	2'
+2.83	1'	7'	5'	3	4	6	2'