XspecT 0.1.2__py3-none-any.whl → 0.1.3__py3-none-any.whl

{XspecT-0.1.2.dist-info → XspecT-0.1.3.dist-info}/METADATA

@@ -1,6 +1,6 @@
 Metadata-Version: 2.1
 Name: XspecT
-Version: 0.1.2
+Version: 0.1.3
 Summary: Tool to monitor and characterize pathogens using Bloom filters.
 License: MIT License
         

{XspecT-0.1.2.dist-info → XspecT-0.1.3.dist-info}/RECORD

@@ -1,16 +1,16 @@
-xspect/BF_v2.py,sha256=3zJgWY6VxfE-6eSqUTgoOY4Z_mp6IKBKDpuWu34FlKI,26080
+xspect/BF_v2.py,sha256=05sp27VzxFtsjt2oyMyhW4aqNmUxGTori49j6lxo1BU,25392
 xspect/Bootstrap.py,sha256=AYyEBo3MoOnPqhPAHe726mX8L9NuXDa5SATxZKLMv3s,830
 xspect/Classifier.py,sha256=BgqpZiMYi2maaccTzJcgH2tjrtDH-U7COc7E4t4cQt8,3602
 xspect/OXA_Table.py,sha256=1GxsyxMpUEgQirY0nJHtR3jl61DoPZh2Rb9L0VdMxD4,1632
-xspect/WebApp.py,sha256=eo1EJOMjW5grCZyvX5g1J4ppwyZb_M9lYGCNuJidM0Q,25224
-xspect/XspecT_mini.py,sha256=t_4OlhzLytRXkM0ig9lo0Szfm2QgJhls52TScUxFN1s,55411
+xspect/WebApp.py,sha256=H4NyfDELrqUSFKOGDLNSJxsNzfLsCX9_BJMln9UXQk0,24941
+xspect/XspecT_mini.py,sha256=OApDXSVIZFK8ZNpNJRPYTlyOLszZbpkJt3jJC51hV8Q,54694
 xspect/XspecT_trainer.py,sha256=6Gj2mltyVyM8Rsh5EU8tSCGMG7niYBLfId664zYaVXI,21703
 xspect/__init__.py,sha256=47DEQpj8HBSa-_TImW-5JCeuQeRkm5NMpJWZG3hSuFU,0
 xspect/download_filters.py,sha256=wSyX-IucjuKIEcVx-E0ClsA0XL0DI1FgMlO2UULgaXc,1048
 xspect/file_io.py,sha256=IWae7xxAt-EmyEbxo0nDSe3RJHmLkQT5jNS2Z3qLKdg,4807
-xspect/main.py,sha256=bF7ntgy_gR0ZNIB9JVxtXb-a6o0Lt0__tI_zzj03B24,2977
+xspect/main.py,sha256=xZxaXOUvA26OurM9DLsbqDzpUEvbtTZOs4916y2Ifdo,2948
 xspect/map_kmers.py,sha256=63iTQS_GZZBK2DxjEs5xoI4KgfpZOntCKul06rrgi5w,6000
-xspect/search_filter.py,sha256=EZkM2917cjy4Q0zQDC9bJ0S-dyD-MBBmJqrAHQ1P260,17190
+xspect/search_filter.py,sha256=a0n2VHmmfVqXaKwLLb478Lvb46kN2GwNgFOZmee1_xo,17261
 xspect/static/How-To.png,sha256=QO6HydIHcL3oM9feMxmfZcKE8M62fIRl2xs_5S_NL5M,119621
 xspect/static/Logo.png,sha256=bvOWMpqxmBigg9jEvZtIMOsXncbSFwnYu4eYNSf1_Qw,296095
 xspect/static/Logo2.png,sha256=V7hpGb3XYLN5vEQQNJdpNjQX_F2A_f1wKAP8N37NwGs,292730
@@ -28,8 +28,9 @@ xspect/templates/layoutabout.html,sha256=ICC8g0DP8a7MLNrEYnXBgtnkwMjIktsimmqwqjM
 xspect/templates/layouthome.html,sha256=6EtVD-L6xlTc7XGk77f9CARKW7JLpv2iiyUci1BK00A,2870
 xspect/templates/layoutspecies.html,sha256=MNGSDEvuKFvgsyXoRLCu-rma10gntUI9vP_9a2sNl7M,24008
 xspect/templates/species.html,sha256=rD9fCmSgyI8hRcmy56mNQH7VR5jnmtriv9WlvTIJJjE,2412
+xspect/train_filter/README_XspecT_Erweiterung.md,sha256=Gn64Biz32LiUfQVYb43Hez6ihDTSFTLfEMwUJ_l1MGU,2879
 xspect/train_filter/__init__.py,sha256=47DEQpj8HBSa-_TImW-5JCeuQeRkm5NMpJWZG3hSuFU,0
-xspect/train_filter/create_svm.py,sha256=E1QwBeUtAlOlKf6QKfmRtKaz_6idv7M8Hb-jbNb_wGk,6820
+xspect/train_filter/create_svm.py,sha256=sktMp8tmQWjcbWkiEzRizabB4qF9wVr7bds2jAvxx4Y,6830
 xspect/train_filter/extract_and_concatenate.py,sha256=kXGqCrOk3TbOkKLJV8nKC6nL8Zg0TWKDCJu2gq8K_cw,5239
 xspect/train_filter/get_paths.py,sha256=JXPbv_Fx5BKHZQ4bkSIGU7yj5zjkmhsI0Z6U4nU0gug,941
 xspect/train_filter/html_scrap.py,sha256=iQXREhG37SNUx7gHoP8eqayMEIH00QLFMTNmIMogb_M,3799
@@ -40,9 +41,9 @@ xspect/train_filter/ncbi_api/download_assemblies.py,sha256=iX1qK8R6p2b3RiHPfqVsL
 xspect/train_filter/ncbi_api/ncbi_assembly_metadata.py,sha256=RhHvxKiQ8HJgoSb6njYEgO_vPioBqEMPvT3lE2lHXp0,3766
 xspect/train_filter/ncbi_api/ncbi_children_tree.py,sha256=pmzg6-fDGLinNSXNbBRv0v62lRgHxW4aXZ0uV1TJhOE,1793
 xspect/train_filter/ncbi_api/ncbi_taxon_metadata.py,sha256=uhBBGffgL4mcJpyp9KxVyOGUh8FxUTAI4xKzoLDav_Y,1577
-XspecT-0.1.2.dist-info/LICENSE,sha256=bhBGDKIRUVwYIHGOGO5hshzuVHyqFJajvSOA3XXOLKI,1094
-XspecT-0.1.2.dist-info/METADATA,sha256=h4OX8L719oZsPj0Xcab4bx4ZstZiMUuPFpVcbZoGc_w,5475
-XspecT-0.1.2.dist-info/WHEEL,sha256=oiQVh_5PnQM0E3gPdiz09WCNmwiHDMaGer_elqB3coM,92
-XspecT-0.1.2.dist-info/entry_points.txt,sha256=L7qliX3pIuwupQxpuOSsrBJCSHYPOPNEzH8KZKQGGUw,43
-XspecT-0.1.2.dist-info/top_level.txt,sha256=hdoa4cnBv6OVzpyhMmyxpJxEydH5n2lDciy8urc1paE,7
-XspecT-0.1.2.dist-info/RECORD,,
+XspecT-0.1.3.dist-info/LICENSE,sha256=bhBGDKIRUVwYIHGOGO5hshzuVHyqFJajvSOA3XXOLKI,1094
+XspecT-0.1.3.dist-info/METADATA,sha256=ulwNLqzESiHPxspAckJm3RkWXg5qf-T6KoNvfTnsH0g,5475
+XspecT-0.1.3.dist-info/WHEEL,sha256=oiQVh_5PnQM0E3gPdiz09WCNmwiHDMaGer_elqB3coM,92
+XspecT-0.1.3.dist-info/entry_points.txt,sha256=L7qliX3pIuwupQxpuOSsrBJCSHYPOPNEzH8KZKQGGUw,43
+XspecT-0.1.3.dist-info/top_level.txt,sha256=hdoa4cnBv6OVzpyhMmyxpJxEydH5n2lDciy8urc1paE,7
+XspecT-0.1.3.dist-info/RECORD,,

xspect/BF_v2.py

@@ -48,6 +48,7 @@ class AbaumanniiBloomfilter:
     ]  # names of the IC's
     number_of_kmeres = 0  # counter of k-meres, will be used to calculate score
     reads = 1000  # standard read number
+    kmer_hits_single = []  # kmer hits per filter
 
     def __init__(self, arraysize):
         """creates empty matrix"""
@@ -181,8 +182,17 @@ class AbaumanniiBloomfilter:
 
         return positions
 
+    def lookup_canonical(self, kmer, limit=False):
+        """takes kmer input string and checks all clonetypes if the cononicalized kmer is inside that set of kmers"""
+
+        # canonicalize
+        complement = str(Seq(kmer).reverse_complement())
+        kmer = max(kmer, complement)
+
+        self.lookup(kmer, limit)
+
     def lookup(self, kmer, limit=False):
-        """checks if an element is in the filters, returns list with True/False,
+        """
         takes kmer input string and checks all clonetypes if the k-mer is inside that set of kmers
         """
 
@@ -213,6 +223,8 @@ class AbaumanniiBloomfilter:
                 temp[i] += 1
                 self.hit = True
                 if limit:
+                    # reset single kmer kit vector / memory management
+                    self.kmer_hits_single = []
                     if self.table.lookup(self.names[i], kmer):
                         self.hits_per_filter[i] += 1
                 else:
@@ -274,11 +286,7 @@ class AbaumanniiBloomfilter:
                         continue
                     self.number_of_kmeres += 1
                     kmer = str(single_read[j : j + self.k])
-                    kmer_reversed = str(Seq(kmer).reverse_complement())
-                    if kmer > kmer_reversed:
-                        self.lookup(kmer)
-                    else:
-                        self.lookup(kmer_reversed)
+                    self.lookup_canonical(kmer)
         # XspecT Sequence-Reads every 10th kmer
         elif quick == 2:
             for single_read in range(0, len(reads)):
@@ -291,11 +299,7 @@ class AbaumanniiBloomfilter:
                     # lookup for kmer
                     temp = reads[single_read]
                     kmer = str(temp[j : j + self.k])
-                    kmer_reversed = str(Seq(kmer).reverse_complement())
-                    if kmer > kmer_reversed:
-                        self.lookup(kmer)
-                    else:
-                        self.lookup(kmer_reversed)
+                    self.lookup_canonical(kmer)
                     if self.hit == True:
                         hit_counter += 1
         elif quick == 3:
@@ -307,11 +311,7 @@ class AbaumanniiBloomfilter:
                         continue
                     self.number_of_kmeres += 1
                     kmer = str(single_read[j : j + self.k])
-                    kmer_reversed = str(Seq(kmer).reverse_complement())
-                    if kmer > kmer_reversed:
-                        self.lookup(kmer)
-                    else:
-                        self.lookup(kmer_reversed)
+                    self.lookup_canonical(kmer)
         # metagenome mode
         elif quick == 4:
             print("Stage 1")
@@ -332,13 +332,8 @@ class AbaumanniiBloomfilter:
                 self.hits_per_filter = [0] * self.clonetypes
                 for kmer in read:
                     counter += 1
-                    # lookup for kmer, use lexikographical smaller kmer
                     self.number_of_kmeres += 1
-                    kmer_reversed = str(Seq(kmer).reverse_complement())
-                    if kmer > kmer_reversed:
-                        self.lookup(kmer)
-                    else:
-                        self.lookup(kmer_reversed)
+                    self.lookup_canonical(kmer)
                 score = self.get_score()
                 score_edit = [str(x) for x in score]
                 score_edit = ",".join(score_edit)
@@ -383,9 +378,9 @@ class AbaumanniiBloomfilter:
                 bootstrap_score = bootstrap_score / bootstrap_n
                 # bootstrap_score = 1
 
-                if ("A." + prediction) not in reads_classified:
+                if prediction not in reads_classified:
                     # Value 5 war vohrer = read
-                    reads_classified["A." + prediction] = [
+                    reads_classified[prediction] = [
                         [max(score)],
                         1,
                         [len(read)],
@@ -395,13 +390,11 @@ class AbaumanniiBloomfilter:
                         None,
                     ]
                 else:
-                    reads_classified["A." + prediction][0] += [max(score)]
-                    reads_classified["A." + prediction][1] += 1
-                    reads_classified["A." + prediction][2] += [len(read)]
-                    reads_classified["A." + prediction][3] += sorted(score)[-2] / max(
-                        score
-                    )
-                    reads_classified["A." + prediction][4] += [bootstrap_score]
+                    reads_classified[prediction][0] += [max(score)]
+                    reads_classified[prediction][1] += 1
+                    reads_classified[prediction][2] += [len(read)]
+                    reads_classified[prediction][3] += sorted(score)[-2] / max(score)
+                    reads_classified[prediction][4] += [bootstrap_score]
                     # reads_classified["A." + prediction][5] += None
                 # tracker.print_diff()
             # not ready yet
@@ -475,11 +468,7 @@ class AbaumanniiBloomfilter:
                     self.number_of_kmeres += 1
                     # lookup for kmer
                     kmer = str(single_read[j : j + self.k])
-                    kmer_reversed = str(Seq(kmer).reverse_complement())
-                    if kmer > kmer_reversed:
-                        self.lookup(kmer)
-                    else:
-                        self.lookup(kmer_reversed)
+                    self.lookup_canonical(kmer)
 
     def cleanup(self):
         """deletes matrix"""

xspect/WebApp.py

@@ -12,7 +12,7 @@ import logging
 import pickle
 import secrets
 import pandas as pd
-from Bio import Entrez, Medline
+from Bio import Entrez, Medline, SeqIO
 from flask import (
     Flask,
     render_template,
@@ -139,35 +139,25 @@ def assignspec():
         return redirect("/resultsspec")
 
     else:
-        # Checking file type
-        # if the file is fasta -> concat lines
-        ext = filename.split(".")[-2]
+        ext = filename.split(".")[-1]
         with open(filename) as f:
             reads = f.read().splitlines()
 
-        # Concat Lines if not .fq file
-        if ext != "fq" and ext != "fastq":
-            reads = "".join(reads)
-            reads = reads.split(">")
+        if ext == "fq" or ext == "fastq":
+            sequences = SeqIO.parse(filename, "fastq")
+            quick = 2
+        else:
             if quick:
                 quick = 1
             else:
                 quick = 0
-            if metagenome:
-                quick = 4
-            reads.pop(0)
-        else:
-            if metagenome:
-                quick = 4
-            else:
-                quick = 2
-        # deleting file
-        os.remove(filename)
+            sequences = SeqIO.parse(filename, "fasta")
+
+        reads = [str(sequence.seq).upper() for sequence in sequences]
 
-        for i in range(len(reads)):
-            reads[i] = reads[i].upper()
     # starts the lookup for a given sequence
     if metagenome:
+        quick = 4
         start_meta = time.time()
         reads, reads_oxa = read_search_pre(reads, BF_Master_prefilter, ext)
         end_meta = time.time()
@@ -193,8 +183,8 @@ def assignspec():
         # assign reads to species
         species_dict = {}
         predictions_names = set()
-        for ele in predictions:
-            predictions_names.add(ele)
+        for prediction in predictions:
+            predictions_names.add(prediction)
         for species in predictions_names:
             species_dict[species] = []
         # dict with species as keys and reads as values for oxa search
@@ -323,11 +313,6 @@ def assignspec():
         )
         return redirect("/resultsspec")
 
-    app.logger.info(
-        "Assignment done for " + str(filename) + ", Time needed: " + str(needed)
-    )
-    return redirect("/resultsspec")
-
 
 # about page
 @app.route("/about")
@@ -511,6 +496,7 @@ def resultsspec():
 
     elif metagenome:
         reads_classified = session.get("reads_classified")
+        genus = session.get("genus")
         # sort reads_classified by highest value of the second element
         sorted_reads_classified = dict(
             sorted(reads_classified.items(), key=lambda x: x[1][1], reverse=True)
@@ -518,8 +504,9 @@ def resultsspec():
         # get key of reads_classified with highest value of the second element from the value
         predictions = []
         values = []
+        translation_dict = load_translation_dict(genus)
         for key, value in sorted_reads_classified.items():
-            predictions.append(key)
+            predictions.append(translation_dict[key])
             values.append(value[1])
         clonetypes_sorted = predictions[:12]
         values_sorted = values[:12]

xspect/XspecT_mini.py

@@ -608,7 +608,9 @@ def xspecT(BF, BF_1_1, files, paths, file_format, read_amount, metagenome, genus
                     threshold_contig = sample_size * 0.7
                     for i in range(0, len(str(sequence.seq)) - BF_1_1.k, sample_size):
                         if "N" not in str(sequence.seq[i : i + BF_1_1.k]):
-                            BF_1_1.lookup(str(sequence.seq[i : i + BF_1_1.k]).upper())
+                            BF_1_1.lookup_canonical(
+                                str(sequence.seq[i : i + BF_1_1.k]).upper()
+                            )
 
                     # needs at least 70% hits to continue with the contig
                     counter = 0
@@ -620,7 +622,7 @@ def xspecT(BF, BF_1_1, files, paths, file_format, read_amount, metagenome, genus
                                 )
                                 counter += 1
                                 # how many kmers? to use
-                                if counter >= 5000000:
+                                if counter >= 5000:
                                     break
                         # contigs_kmers.append(str(reverse_sequence[j: j + BF_1_1.k]))
                         contigs.append(contigs_kmers)
@@ -638,7 +640,7 @@ def xspecT(BF, BF_1_1, files, paths, file_format, read_amount, metagenome, genus
                         for j in range(len(contigs[i])):
                             BF_1_1.number_of_kmeres += 1
                             hits_per_filter_copy = BF_1_1.hits_per_filter[:]
-                            BF_1_1.lookup(contigs[i][j])
+                            BF_1_1.lookup_canonical(contigs[i][j])
                             if hits_per_filter_copy != BF_1_1.hits_per_filter:
                                 threshold += 1
                         # parameter value needs to be determined
@@ -654,11 +656,7 @@ def xspecT(BF, BF_1_1, files, paths, file_format, read_amount, metagenome, genus
                     BF.hits_per_filter = [0] * BF.clonetypes
                     for kmer in contigs_filtered:
                         BF.number_of_kmeres += 1
-                        kmer_reversed = str(Seq.Seq(kmer).reverse_complement())
-                        if kmer > kmer_reversed:
-                            BF.lookup(kmer)
-                        else:
-                            BF.lookup(kmer_reversed)
+                        BF.lookup_canonical(kmer)
                     score = BF.get_score()
                     score_edit = [str(x) for x in score]
                     score_edit = ",".join(score_edit)
@@ -750,11 +748,7 @@ def xspecT(BF, BF_1_1, files, paths, file_format, read_amount, metagenome, genus
                     for j in range(0, len(sequence.seq) - BF.k, mode):
                         BF.number_of_kmeres += 1
                         kmer = str(sequence.seq[j : j + BF.k])
-                        kmer_reversed = str(Seq.Seq(kmer).reverse_complement())
-                        if kmer > kmer_reversed:
-                            BF.lookup(kmer)
-                        else:
-                            BF.lookup(kmer_reversed)
+                        BF.lookup_canonical(kmer)
 
             score = BF.get_score()
             # print("Scores: ", score)
@@ -884,11 +878,11 @@ def xspecT(BF, BF_1_1, files, paths, file_format, read_amount, metagenome, genus
                     #    if "N" not in str(sequence.seq[i: i + BF_1_1.k]):
                     #        BF_1_1.lookup(str(sequence.seq[i: i + BF_1_1.k]))
                     if "N" not in str(sequence.seq):
-                        BF_1_1.lookup(k1)
-                        BF_1_1.lookup(k2)
-                        BF_1_1.lookup(k3)
-                        BF_1_1.lookup(k4)
-                        BF_1_1.lookup(k5)
+                        BF_1_1.lookup_canonical(k1)
+                        BF_1_1.lookup_canonical(k2)
+                        BF_1_1.lookup_canonical(k3)
+                        BF_1_1.lookup_canonical(k4)
+                        BF_1_1.lookup_canonical(k5)
                     else:
                         continue
                     # needs at least 2 of 3 hits to continue with read
@@ -913,7 +907,7 @@ def xspecT(BF, BF_1_1, files, paths, file_format, read_amount, metagenome, genus
                             BF_1_1.number_of_kmeres += 1
                             hits_per_filter_copy = BF_1_1.hits_per_filter[:]
                             if "N" not in reads[i][j]:
-                                BF_1_1.lookup(reads[i][j])
+                                BF_1_1.lookup_canonical(reads[i][j])
                             if hits_per_filter_copy != BF_1_1.hits_per_filter:
                                 threshold += 1
                         if threshold >= 0.7 * len(reads[i]):
@@ -929,11 +923,7 @@ def xspecT(BF, BF_1_1, files, paths, file_format, read_amount, metagenome, genus
                     for kmer in reads_filtered:
                         if "N" not in kmer:
                             BF.number_of_kmeres += 1
-                            kmer_reversed = str(Seq.Seq(kmer).reverse_complement())
-                            if kmer > kmer_reversed:
-                                BF.lookup(kmer)
-                            else:
-                                BF.lookup(kmer_reversed)
+                            BF.lookup_canonical(kmer)
                         else:
                             continue
                     score = BF.get_score()
@@ -1041,11 +1031,7 @@ def xspecT(BF, BF_1_1, files, paths, file_format, read_amount, metagenome, genus
                         for j in range(0, len(sequence.seq) - BF.k + 1, mode):
                             BF.number_of_kmeres += 1
                             kmer = str(sequence.seq[j : j + BF.k])
-                            kmer_reversed = str(Seq.Seq(kmer).reverse_complement())
-                            if kmer > kmer_reversed:
-                                BF.lookup(kmer)
-                            else:
-                                BF.lookup(kmer_reversed)
+                            BF.lookup_canonical(kmer)
                     else:
                         break
             score = BF.get_score()

xspect/main.py

@@ -56,6 +56,7 @@ def classify(genus, path, species, ic, oxa, metagenome, complete, save):
         mode = 1
     file_format = "fasta"
     read_amount = 342480
+    print(mode)
 
     xspecT_mini(
         path,
@@ -111,7 +112,6 @@ def train(genus, bf_assembly_path, svm_assembly_path, complete, check):
 @cli.command()
 def web():
     """Open the XspecT web app."""
-    webbrowser.open("http://localhost:8000")
     app.run(host="0.0.0.0", port=8000, debug=True, threaded=True)
 
 

xspect/search_filter.py

@@ -299,7 +299,7 @@ def read_search_pre(reads, BF_pre, ext):
             threshold_read = sample_size * 0.7
             for i in range(0, len(single_read) - BF_pre.k, sample_size):
                 if "N" not in single_read[i : i + BF_pre.k]:
-                    BF_pre.lookup(single_read[i : i + BF_pre.k])
+                    BF_pre.lookup_canonical(single_read[i : i + BF_pre.k])
         # for reads use a static sample of 5
         # Taking sum of list as reference, if sum has not increased after testing those 3 kmeres,
         # then the read won't be tested further
@@ -312,11 +312,11 @@ def read_search_pre(reads, BF_pre, ext):
             k4 = single_read[BF_pre.k : BF_pre.k * 2]
             k5 = single_read[mid + BF_pre.k : mid + BF_pre.k * 2]
             if "N" not in single_read:
-                BF_pre.lookup(k1)
-                BF_pre.lookup(k2)
-                BF_pre.lookup(k3)
-                BF_pre.lookup(k4)
-                BF_pre.lookup(k5)
+                BF_pre.lookup_canonical(k1)
+                BF_pre.lookup_canonical(k2)
+                BF_pre.lookup_canonical(k3)
+                BF_pre.lookup_canonical(k4)
+                BF_pre.lookup_canonical(k5)
             threshold_read = 3
         # needs at least 2 of 3 hits to continue with read
         counter = 0
@@ -348,7 +348,7 @@ def read_search_pre(reads, BF_pre, ext):
         for j in range(len(reads_new[i])):
             BF_pre.number_of_kmeres += 1
             hits_per_filter_copy = BF_pre.hits_per_filter[:]
-            BF_pre.lookup(reads_new[i][j])
+            BF_pre.lookup_canonical(reads_new[i][j])
             if hits_per_filter_copy != BF_pre.hits_per_filter:
                 threshold += 1
         if threshold >= cutoff * len(reads_new[i]):
@@ -373,7 +373,7 @@ def read_search_spec(reads, quick, BF, ext, genus):
         names = [translation_dict[name] for name in names_id]
         return score, names, hits, None
     # Metagenome mode
-    elif quick == 4:
+    elif  quick == 4:
         reads_classified, predictions = BF.lookup_txt(reads, genus, ext, quick)
         hits = None
         names = None

xspect/train_filter/README_XspecT_Erweiterung.md

@@ -0,0 +1,119 @@
+# XspecT-Erweiterung
+
+Expands XspecT, so new filter for a genus can automatically be trained. It's main 
+script is XspecT_trainer.py. The rest of the scripts are inside the python module 
+train_filter. 
+
+## Training new filter
+
+XspecT_trainer.py uses command line arguments. The examples for using XspecT_trainer.py
+are using Salmonella since this genus only has two defined species in the NCBI 
+databases.
+
+### Jellyfish
+
+The program jellyfish is used to count distinct k-meres in the assemblies. For XspecT_
+trainer.py to work jellyfish needs to be installed. It can be installed using bioconda:
+
+`
+conda install -c bioconda jellyfish
+`
+
+### Training examples
+
+New filters with assemblies from NCBI RefSeq can be trained with the following line. The 
+python libraries from [requirements.txt](..%2Frequirements.txt) need to be installed.
+
+`
+python XspecT_trainer.py Salmonella 1
+`
+
+Training filters with custom data can be done using the following line.
+
+`
+python XspecT_trainer.py Salmonella 2 -bf /path/to/concate_assemblies -svm 
+/path/to/assemblies
+`
+
+All command line arguments are explained using the following line.
+
+`
+python XspecT_trainer.py -h
+`
+
+# Explanation of the scripts
+
+## backup_filter.py
+
+Creates a backup of all files needed for the species assignment by XspecT for a specific
+genus. The backup will be done, if new filters will be created for a genus which 
+already has trained filters.
+
+## create_svm.py
+
+Downloads the needed assemblies and trains a support-vector-machine for the genus.
+
+## extract_and_concatenate.py
+
+Unzips the downloaded assemblies. Concatenates assemblies per species that will be used 
+to train the bloomfilters.
+
+## get_paths.py
+
+Functions that get specific paths.
+
+## html_scrap.py
+
+Updates a list of all NCBI RefSeq assembly accessions that have a taxonomy check result
+of OK. The taxonomy check from NCBI RefSeq uses the ANI (average-nucleotide-
+identity) to compute a result.
+
+## interface_XspecT.py
+
+Mostly functions that train new bloomfilters automatically. The functions were 
+originally writen for XspecT in a non-automatic way and were updated.
+
+## k_mer_count.py
+
+Uses jellyfish to count distinct k-meres in every concatenated assembly. The highest
+count will be used to compute the size of the bloomfilters.
+
+## ncbi_api
+
+A module which makes requests to the NCBI Datasets API.
+
+### download_assemblies.py
+
+The specific function that downloads assemblies from NCBI RefSeq using NCBI 
+datasets.
+
+### ncbi_assembly_metadata.py
+
+Takes a dictionary with species and their taxon ID and asks NCBI for assemblies of
+the species. Saves the collected accessions of the found and selected assemblies.
+
+### ncbi_children_tree.py
+
+Takes the name or ID of a genus and gives a list with all its species.
+
+### ncbi_taxon_metadata.py
+
+Takes a list with taxon and collects metadata like their scientific name and rank.
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+

xspect/train_filter/create_svm.py

@@ -127,7 +127,7 @@ def perform_lookup(bloomfilter, files, file_paths, accessions, names, spacing):
             # Dominik: changed sample size to var
             for j in range(0, len(sequence.seq) - BF.k, spacing):
                 BF.number_of_kmeres += 1
-                BF.lookup(str(sequence.seq[j : j + BF.k]))
+                BF.lookup_canonical(str(sequence.seq[j : j + BF.k]))
 
         score = BF.get_score()
         score = [str(x) for x in score]