geney 1.3.60__py2.py3-none-any.whl → 1.3.62__py2.py3-none-any.whl

geney/SeqMats.py

@@ -358,6 +358,7 @@ class SeqMat:
         return SeqMat(ref_seqmat)
 
     def orf_seqmat(self, tis_index):
+        stop_index = None
         if tis_index not in self.indices:
             return SeqMat.from_seq({'seq': ''})
 
@@ -367,21 +368,17 @@ class SeqMat:
         # temp = temp[:, temp[0, :] != 5]
         # temp = SeqMat(temp)  # .drop_indices()
         # raw_seq = temp.seq  # Extract the raw sequence
-        pattern = re.compile(r"(?:[NACGT]{3})*?(TAA|TAG|TGA)")
-        match = pattern.match(raw_seq.seq)
-        print(match)
-        # match = [(m.group(), m.start()+1) for m in re.finditer(r"(?:[NACGT]{3})*?(TAA|TAG|TGA)", raw_seq.seq)]# if
-        #          # (m.start()+1) % 3 == 0]
-
-        # print(match)
-        if len(match) > 0:
-            stop_index = match[0][-1] #.end()
-        else:
+        pattern = re.compile(r"TAA|TAG|TGA")
+        matches = pattern.finditer(raw_seq.seq)  # Use finditer to get all matches
+        for m in matches:
+            if m.start() % 3 == 0:
+                stop_index = m.start()
+                break
+
+        if stop_index is None:
             stop_index = len(raw_seq) - (len(raw_seq) % 3)
 
-        end_index = stop_index
-        # assert (end_index+1) % 3 == 0, f"{end_index} is not a multiple of 3"
-        return SeqMat(raw_seq.seqmat[:, :end_index])
+        return SeqMat(raw_seq.seqmat[:, :stop_index])
 
     def translate(self, tis_index):
         """

geney/oncosplice.py

@@ -369,6 +369,7 @@ def oncosplice(mut_id, splicing_threshold=0.5, protein_coding=True, cons_require
                     splicing_db.store_mutation_data(engine=splicing_engine, mut_id=mut_id, gene=gene.gene_name, transcript_id=reference_transcript.transcript_id, data=missplicing.missplicing)
 
         alternative_splicing_paths = develop_aberrant_splicing(reference_transcript, missplicing.missplicing)
+        print("iterating through new boundaries...")
         for i, new_boundaries in enumerate(alternative_splicing_paths):
             mutated_transcript.acceptors = new_boundaries['acceptors']
             mutated_transcript.donors = new_boundaries['donors']
@@ -413,6 +414,7 @@ def oncosplice(mut_id, splicing_threshold=0.5, protein_coding=True, cons_require
             results.append(report)
 
     if len(results) == 0:
+        print("Nothing...")
         return None
 
     return pd.DataFrame(results)[

geney/utils.py

@@ -78,3 +78,24 @@ def is_monotonic(A):
 #     return rev_compl
 #
 
+
+def generate_random_nucleotide_sequences(num_sequences, min_len=3, max_len=10):
+    """
+    Generate random sequences of nucleotides.
+
+    Parameters:
+        num_sequences (int): Number of sequences to generate.
+        sequence_length (int): Length of each sequence.
+
+    Returns:
+        list: A list of random nucleotide sequences.
+    """
+    import random
+    nucleotides = ['A', 'C', 'G', 'T']
+    lengths = list(range(min_len, max_len))
+    sequences = [
+        ''.join(random.choices(nucleotides, k=random.choice(lengths)))
+        for _ in range(num_sequences)
+    ]
+    return sequences
+

{geney-1.3.60.dist-info → geney-1.3.62.dist-info}/METADATA

@@ -1,6 +1,6 @@
 Metadata-Version: 2.1
 Name: geney
-Version: 1.3.60
+Version: 1.3.62
 Summary: A Python package for gene expression modeling.
 Home-page: https://github.com/nicolaslynn/geney
 Author: Nicolas Lynn

{geney-1.3.60.dist-info → geney-1.3.62.dist-info}/RECORD

@@ -1,6 +1,6 @@
 geney/Fasta_segment.py,sha256=99HxNGNh_MfdVW6hhtlb1vOn7eSmT7oFoEfHDFMxG8w,11275
 geney/Gene.py,sha256=nMWJjoQaiVFm2iRjoiq7ghZqnXtW0tJDcq2S0AyOIvY,6883
-geney/SeqMats.py,sha256=DwEq_452FsF-zrr2nE1MMSTfKf54931yJERfXq0Ak7k,19460
+geney/SeqMats.py,sha256=ksS1JW3vDj-CRBpcy-K28UFZSfJaMuc-JNNM26T09us,19259
 geney/Transcript.py,sha256=CpfxYkuCwFILozrtLuiWnlr1mRnMKn4o84HVJislgYs,14499
 geney/__init__.py,sha256=eBdDl42N6UhcYeZDjOnv199Z88fI5_8Y6xW8447OKXM,755
 geney/_mutation_utils.py,sha256=dHssUsnii_mf-wuRoMmF13UlD7k3ml_VwQMItTYnXpU,1132
@@ -11,7 +11,7 @@ geney/graphic_utils.py,sha256=oMsBpB9YeEn96gGpKh4MmtagJffWZbk-xPrIwHvkFhA,11016
 geney/gtex_utils.py,sha256=asL2lHyU5KsbWpV096vkf1Ka7hSo_RRfZqw7p5nERmE,1919
 geney/immune_utils.py,sha256=ZRni5ttrhpYBnmNr0d0ZatIbNPYs4nmQuoUO00SpsS4,5271
 geney/mutation_utils.py,sha256=C_kv2MB_L8LlhX3W2ooXjJ3uDoJ8zX1WeDtZKoBZJkI,1547
-geney/oncosplice.py,sha256=9_Eo5fL68CosTj6yTfh3UjCXHprhzqHFRL21WRicMRs,24704
+geney/oncosplice.py,sha256=YyoJEBY8yKsaKGFXmH_nJRfzb6gBfuMq9lV89H3A9D0,24785
 geney/pangolin_utils.py,sha256=9jdBXlOcRaUdfi-UpUxHA0AkTMZkUF-Lt7HVZ1nEm3s,2973
 geney/power_utils.py,sha256=MehZFUdkJ2EFUot709yPEDxSkXmH5XevMebX2HD768A,7330
 geney/seqmat_utils.py,sha256=wzb3PX5it5bpIFQvcxyzlxfhoJTbHHbsjg0rzh05iVs,19753
@@ -20,12 +20,12 @@ geney/splicing_utils.py,sha256=GDSuuy8kWBh2_4UDwqjclF9gKnnCoLc_neOEsSrNWQM,38512
 geney/survival_utils.py,sha256=KnAzEviMuXh6SnVXId9PgsFLSbgkduTvYoIthxN7FPA,6886
 geney/tcga_utils.py,sha256=D_BNHm-D_K408dlcJm3hzH2c6QNFjQsKvUcOPiQRk7g,17612
 geney/tis_utils.py,sha256=la0CZroaKe5RgAyFd4Bf_DqQncklWgAY2823xVst98o,7813
-geney/utils.py,sha256=EsKvBM-Nz2a3_4ZAhF4Dxd4PwT7_6YYKpxEN4LLgg10,2174
+geney/utils.py,sha256=KBdwNIywo7INVEQEsuIXauEJobvReE9TXAi5qqXanSI,2775
 geney/translation_initiation/__init__.py,sha256=47DEQpj8HBSa-_TImW-5JCeuQeRkm5NMpJWZG3hSuFU,0
 geney/translation_initiation/tis_utils.py,sha256=AF3siFjuQH-Rs44EV-80zHdbxRMvN4woLFSHroWIETc,4448
 geney/translation_initiation/resources/kozak_pssm.json,sha256=pcd0Olziutq-6H3mFWDCD9cujQ_AlZO-iiOvBl82hqE,1165
 geney/translation_initiation/resources/tis_regressor_model.joblib,sha256=IXb4DUDhJ5rBDKcqMk9zE3ECTZZcdj7Jixz3KpoZ7OA,2592025
-geney-1.3.60.dist-info/METADATA,sha256=9RdNxrkkxfNmokUDoatZq-ALXfGVcfyXXR5bEFI0_Ck,990
-geney-1.3.60.dist-info/WHEEL,sha256=AHX6tWk3qWuce7vKLrj7lnulVHEdWoltgauo8bgCXgU,109
-geney-1.3.60.dist-info/top_level.txt,sha256=O-FuNUMb5fn9dhZ-dYCgF0aZtfi1EslMstnzhc5IIVo,6
-geney-1.3.60.dist-info/RECORD,,
+geney-1.3.62.dist-info/METADATA,sha256=JaEKHBPqwIUagmuVLomjmNJhs0vF_pT3bK6z7gsPGRc,990
+geney-1.3.62.dist-info/WHEEL,sha256=AHX6tWk3qWuce7vKLrj7lnulVHEdWoltgauo8bgCXgU,109
+geney-1.3.62.dist-info/top_level.txt,sha256=O-FuNUMb5fn9dhZ-dYCgF0aZtfi1EslMstnzhc5IIVo,6
+geney-1.3.62.dist-info/RECORD,,