geney 1.3.79__py2.py3-none-any.whl → 1.4.0__py2.py3-none-any.whl

geney/{oncosplice.py → _oncosplice.py}

@@ -1,22 +1,17 @@
 from Bio import pairwise2
 import re
-import hashlib
 from datetime import datetime
 from tqdm import tqdm
 import pandas as pd
 import numpy as np
-from .SeqMats import SeqMat, MutSeqMat
-from .splicing_utils import find_transcript_missplicing_seqs, develop_aberrant_splicing, Missplicing
+from geney.utils.SeqMats import MutSeqMat
+from ._splicing_utils import find_transcript_missplicing_seqs, develop_aberrant_splicing, Missplicing
 from .Gene import Gene
-import copy
-from . import config
 
-from .tis_utils import find_tis
 
-def short_hash_of_list(numbers, length=5):
-    encoded = repr(numbers).encode('utf-8')
-    full_hash = hashlib.sha256(encoded).hexdigest()
-    return full_hash[:length]
+##############################################################################################################################
+######################################## ONCOSPLICE CALCULATIONS #############################################################
+##############################################################################################################################
 
 def find_continuous_gaps(sequence):
     """Find continuous gap sequences in an alignment."""
@@ -221,6 +216,23 @@ def moving_average_conv(vector, window_size, factor=1):
 
     return np.convolve(vector, np.ones(window_size), mode='same') / window_size
 
+
+def oncosplice_score(reference_protein, variant_protein, conservation_vector, window_length=13):
+    alignment = get_logical_alignment(reference_protein, variant_protein)
+    deleted, inserted = find_indels_with_mismatches_as_deletions(alignment.seqA, alignment.seqB)
+    modified_positions = find_modified_positions(len(reference_protein), deleted, inserted)
+    temp_cons = np.convolve(conservation_vector * modified_positions,
+                            np.ones(window_length)) / window_length
+    percentile = (
+            sorted(conservation_vector).index(
+                next(x for x in sorted(conservation_vector) if x >= max(temp_cons))) / len(
+        conservation_vector))
+    return max(temp_cons), percentile
+
+
+##############################################################################################################################
+#################################### ANNOTATION FUNCTIONS ####################################################################
+##############################################################################################################################
 def find_splice_site_proximity(pos, transcript):
     for i, (ex_start, ex_end) in enumerate(transcript.exons):
         if min(ex_start, ex_end) <= pos <= max(ex_start, ex_end):
@@ -231,7 +243,6 @@ def find_splice_site_proximity(pos, transcript):
             return None, i + 1, abs(pos - in_end), abs(pos - in_start)
 
     return None, None, np.inf, np.inf
-
 def define_missplicing_events(ref, var):
     ref_introns, ref_exons = ref.introns, ref.exons
     var_introns, var_exons = var.introns, var.exons
@@ -300,149 +311,166 @@ def missense_effect(r, v):
                 nt_changes += f'{r.orf.seq[p:p+3]}{p}{v.orf.seq[p:p+3]},'
     return aa_changes, nt_changes
 
-# Annotating
-def OncospliceAnnotator(reference_transcript, variant_transcript, mut, ref_attributes=[], var_attributes=[]):
-    affected_exon, affected_intron, distance_from_5, distance_from_3 = find_splice_site_proximity(np.floor(mut.indices[0]),
-                                                                                                  reference_transcript)
-
-    report = {}
-    report['primary_transcript'] = reference_transcript.primary_transcript
-    report['transcript_id'] = reference_transcript.transcript_id
-    report['reference_protein'] = reference_transcript.protein
-    report['variant_protein'] = variant_transcript.protein
-    report['variant_protein_length'] = len(variant_transcript.protein)
-    descriptions = define_missplicing_events(reference_transcript, variant_transcript)
-    report['exon_changes'] = '|'.join([v for v in descriptions if v])
-    report['splicing_codes'] = summarize_missplicing_event(*descriptions)
-    report['affected_exon'] = affected_exon
-    report['affected_intron'] = affected_intron
-    report['mutation_distance_from_5'] = distance_from_5
-    report['mutation_distance_from_3'] = distance_from_3
-    aa_c, c_c = missense_effect(reference_transcript, variant_transcript)
-    report['missense_effect'] = aa_c
-    report['codon_change'] = c_c
-    return report
 
+def oncosplice_score(reference_protein, variant_protein, conservation_vector, window_length=13):
+    alignment = get_logical_alignment(reference_protein, variant_protein)
+    deleted, inserted = find_indels_with_mismatches_as_deletions(alignment.seqA, alignment.seqB)
+    modified_positions = find_modified_positions(len(reference_protein), deleted, inserted)
+    temp_cons = np.convolve(conservation_vector * modified_positions,
+                            np.ones(window_length)) / window_length
+    percentile = (
+            sorted(conservation_vector).index(
+                next(x for x in sorted(conservation_vector) if x >= max(temp_cons))) / len(
+        conservation_vector))
+    return max(temp_cons), percentile
 
-def oncosplice(mut_id, splicing_threshold=0.5, protein_coding=True, primary_transcript=False,
-               window_length=13, organism='hg38', splicing_engine=None, splicing_db=None, verbose=False,
-               tis_engine=None, target_transcripts=None):
 
-    gene = Gene.from_file(mut_id.split(':')[0], organism=organism)
-    reference_gene_proteins = {
-        transcript.generate_pre_mrna().generate_mature_mrna().generate_protein().protein: transcript.transcript_id for
-        transcript in gene if transcript.transcript_biotype == 'protein_coding'}
-
-    mutations = [MutSeqMat.from_mutid(m) for m in mut_id.split('|')]
-    if gene.rev:
-        mutations = [m.reverse_complement() for m in mutations[::-1]]
-
-    results = []
-    for reference_transcript in tqdm(gene, desc=f'Processing {mut_id}...'):
-        if target_transcripts is not None and reference_transcript.transcript_id not in target_transcripts:
-            continue
-
-        # if (cons_required and not reference_transcript.cons_available) or (
-        #         protein_coding and not reference_transcript.transcript_biotype == 'protein_coding'):
-        if protein_coding and not reference_transcript.transcript_biotype == 'protein_coding':
-            print("Not protein coding...")
-            continue
-
-        current_mutations = [m for m in mutations if m in reference_transcript]
-        if len(current_mutations) == 0:
-            print(f"No mutations within transcript ({reference_transcript.transcript_start} > {reference_transcript.transcript_end})...")
-            continue
-
-        center = np.mean([m.indices[0] for m in current_mutations]) // 1
-
-        mutated_transcript = reference_transcript.clone()
-
-        for mutation in current_mutations:
-            mutated_transcript.mutate(mutation, inplace=True)
-
-        reference_transcript.generate_mature_mrna().generate_protein()
-
-        if len(reference_transcript.protein) < window_length:
-            print(f"> Window length issue {reference_transcript.transcript_id}")
-            continue
-
-        reference_transcript.cons_vector = transform_conservation_vector(reference_transcript.cons_vector,
-                                                                         window=window_length)
-
-        assert len(reference_transcript.protein) == len(
-            reference_transcript.cons_vector), f"Protein ({len(reference_transcript.protein)}) and conservation vector ({len(reference_transcript.cons_vector)}) must be same length."
-
-        if splicing_engine is None:
-            missplicing = Missplicing()
-        else:
-            missplicing, no_splicing_record = None, True
-            if splicing_db is not None:
-                missplicing = Missplicing(splicing_db.get_mutation_data(engine=splicing_engine, mut_id=mut_id, gene=gene.gene_name, transcript_id=reference_transcript.transcript_id))
-                no_splicing_record = missplicing is None
-
-            if missplicing is None:
-                missplicing = find_transcript_missplicing_seqs(
-                    reference_transcript.pre_mrna.get_context(center, context=7500, padding='N'),
-                    mutated_transcript.pre_mrna.get_context(center, context=7500, padding='N'), reference_transcript.donors,
-                    reference_transcript.acceptors, threshold=splicing_threshold, engine=splicing_engine)
-                if no_splicing_record and splicing_db is not None:
-                    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)
-        for i, new_boundaries in enumerate(alternative_splicing_paths):
-            print("iterating through new boundaries...")
-
-            mutated_transcript.acceptors = new_boundaries['acceptors']
-            mutated_transcript.donors = new_boundaries['donors']
-            mutated_transcript.generate_mature_mrna().generate_protein()
-
-            ### Experimental
-            # mutated_transcript.generate_mature_mrna()
-            # if tis_engine is None:
-            #     tis_candidates = [(mutated_transcript.tis, 1)]
-            # else:
-            #     from tis_utils import tis_predictor
-            #     tis_candidates = tis_predictor(mutated_transcript.mature_mrna)
-            #
-            # for tis_candidate, tis_score in tis_candidates:
-            #     mutated_transcript.generate_protein(tis_candidate)
-            ######
-
-            alignment = get_logical_alignment(reference_transcript.protein, mutated_transcript.protein)
-            deleted, inserted = find_indels_with_mismatches_as_deletions(alignment.seqA, alignment.seqB)
-            modified_positions = find_modified_positions(len(reference_transcript.protein), deleted, inserted)
-            temp_cons = np.convolve(reference_transcript.cons_vector * modified_positions,
-                                    np.ones(window_length)) / window_length
-            affected_cons_scores = max(temp_cons)
-            percentile = (
-                    sorted(reference_transcript.cons_vector).index(
-                        next(x for x in sorted(reference_transcript.cons_vector) if x >= affected_cons_scores)) / len(
-                reference_transcript.cons_vector))
-
-            report = OncospliceAnnotator(reference_transcript, mutated_transcript, current_mutations[0])
-            report['mut_id'] = mut_id
-            report['splicing_engine'] = splicing_engine if splicing_engine is not None else 'None'
-            # report['tis_engine'] = tis_engine if tis_engine is not None else 'None'
-            # report['tis_pos'] = tis_candidate
-            # report['tis_score'] = tis_score
-            report['oncosplice_score'] = affected_cons_scores
-            report['percentile'] = percentile
-            report['isoform_id'] = short_hash_of_list(mutated_transcript.exons)
-            report['isoform_prevalence'] = new_boundaries['path_weight']
-            report['full_missplicing'] = missplicing.aberrant_splicing
-            report['missplicing'] = missplicing.max_delta
-            report['reference_resemblance'] = reference_gene_proteins.get(mutated_transcript.protein, None)
-            results.append(report)
-
-    if len(results) == 0:
-        # print("Nothing...")
-        return pd.DataFrame()
-
-    return pd.DataFrame(results)[
-        ['mut_id', 'transcript_id', 'isoform_id', 'primary_transcript', 'missplicing', 'full_missplicing',
-         'exon_changes', 'splicing_codes', 'affected_exon', 'affected_intron', 'mutation_distance_from_5',
-         'mutation_distance_from_3', 'missense_effect', 'codon_change', 'missense_position', 'reference_resemblance',
-         'oncosplice_score', 'percentile', 'isoform_prevalence', 'reference_protein', 'variant_protein', 'splicing_engine']]
+# # Annotating
+# def OncospliceAnnotator(reference_transcript, variant_transcript, mut, ref_attributes=[], var_attributes=[]):
+#     affected_exon, affected_intron, distance_from_5, distance_from_3 = find_splice_site_proximity(np.floor(mut.indices[0]),
+#                                                                                                   reference_transcript)
+#
+#     report = {}
+#     report['primary_transcript'] = reference_transcript.primary_transcript
+#     report['transcript_id'] = reference_transcript.transcript_id
+#     report['reference_protein'] = reference_transcript.protein
+#     report['variant_protein'] = variant_transcript.protein
+#     report['variant_protein_length'] = len(variant_transcript.protein)
+#     descriptions = define_missplicing_events(reference_transcript, variant_transcript)
+#     report['exon_changes'] = '|'.join([v for v in descriptions if v])
+#     report['splicing_codes'] = summarize_missplicing_event(*descriptions)
+#     report['affected_exon'] = affected_exon
+#     report['affected_intron'] = affected_intron
+#     report['mutation_distance_from_5'] = distance_from_5
+#     report['mutation_distance_from_3'] = distance_from_3
+#     aa_c, c_c = missense_effect(reference_transcript, variant_transcript)
+#     report['missense_effect'] = aa_c
+#     report['codon_change'] = c_c
+#     return report
+#
+# ##############################################################################################################################
+# ######################################## ONCOSPLICE PIPELINES ################################################################
+# ##############################################################################################################################
+#
+# def oncosplice(mut_id, splicing_threshold=0.5, protein_coding=True, primary_transcript=False,
+#                window_length=13, organism='hg38', splicing_engine=None, splicing_db=None, verbose=False,
+#                tis_engine=None, target_transcripts=None):
+#
+#     gene = Gene.from_file(mut_id.split(':')[0], organism=organism)
+#     reference_gene_proteins = {
+#         transcript.generate_pre_mrna().generate_mature_mrna().generate_protein().protein: transcript.transcript_id for
+#         transcript in gene if transcript.transcript_biotype == 'protein_coding'}
+#
+#     mutations = [MutSeqMat.from_mutid(m) for m in mut_id.split('|')]
+#     if gene.rev:
+#         mutations = [m.reverse_complement() for m in mutations[::-1]]
+#
+#     results = []
+#     for reference_transcript in tqdm(gene, desc=f'Processing {mut_id}...'):
+#         if target_transcripts is not None and reference_transcript.transcript_id not in target_transcripts:
+#             continue
+#
+#         # if (cons_required and not reference_transcript.cons_available) or (
+#         #         protein_coding and not reference_transcript.transcript_biotype == 'protein_coding'):
+#         if protein_coding and not reference_transcript.transcript_biotype == 'protein_coding':
+#             print("Not protein coding...")
+#             continue
+#
+#         current_mutations = [m for m in mutations if m in reference_transcript]
+#         if len(current_mutations) == 0:
+#             print(f"No mutations within transcript ({reference_transcript.transcript_start} > {reference_transcript.transcript_end})...")
+#             continue
+#
+#         center = np.mean([m.indices[0] for m in current_mutations]) // 1
+#
+#         mutated_transcript = reference_transcript.clone()
+#
+#         for mutation in current_mutations:
+#             mutated_transcript.mutate(mutation, inplace=True)
+#
+#         reference_transcript.generate_mature_mrna().generate_protein()
+#
+#         if len(reference_transcript.protein) < window_length:
+#             print(f"> Window length issue {reference_transcript.transcript_id}")
+#             continue
+#
+#         reference_transcript.cons_vector = transform_conservation_vector(reference_transcript.cons_vector,
+#                                                                          window=window_length)
+#
+#         assert len(reference_transcript.protein) == len(
+#             reference_transcript.cons_vector), f"Protein ({len(reference_transcript.protein)}) and conservation vector ({len(reference_transcript.cons_vector)}) must be same length."
+#
+#         if splicing_engine is None:
+#             missplicing = Missplicing()
+#         else:
+#             missplicing, no_splicing_record = None, True
+#             if splicing_db is not None:
+#                 missplicing = Missplicing(splicing_db.get_mutation_data(engine=splicing_engine, mut_id=mut_id, gene=gene.gene_name, transcript_id=reference_transcript.transcript_id))
+#                 no_splicing_record = missplicing is None
+#
+#             if missplicing is None:
+#                 missplicing = find_transcript_missplicing_seqs(
+#                     reference_transcript.pre_mrna.get_context(center, context=7500, padding='N'),
+#                     mutated_transcript.pre_mrna.get_context(center, context=7500, padding='N'), reference_transcript.donors,
+#                     reference_transcript.acceptors, threshold=splicing_threshold, engine=splicing_engine)
+#                 if no_splicing_record and splicing_db is not None:
+#                     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)
+#         for i, new_boundaries in enumerate(alternative_splicing_paths):
+#             print("iterating through new boundaries...")
+#
+#             mutated_transcript.acceptors = new_boundaries['acceptors']
+#             mutated_transcript.donors = new_boundaries['donors']
+#             mutated_transcript.generate_mature_mrna().generate_protein()
+#
+#             ### Experimental
+#             # mutated_transcript.generate_mature_mrna()
+#             # if tis_engine is None:
+#             #     tis_candidates = [(mutated_transcript.tis, 1)]
+#             # else:
+#             #     from tis_utils import tis_predictor
+#             #     tis_candidates = tis_predictor(mutated_transcript.mature_mrna)
+#             #
+#             # for tis_candidate, tis_score in tis_candidates:
+#             #     mutated_transcript.generate_protein(tis_candidate)
+#             ######
+#
+#             alignment = get_logical_alignment(reference_transcript.protein, mutated_transcript.protein)
+#             deleted, inserted = find_indels_with_mismatches_as_deletions(alignment.seqA, alignment.seqB)
+#             modified_positions = find_modified_positions(len(reference_transcript.protein), deleted, inserted)
+#             temp_cons = np.convolve(reference_transcript.cons_vector * modified_positions,
+#                                     np.ones(window_length)) / window_length
+#             affected_cons_scores = max(temp_cons)
+#             percentile = (
+#                     sorted(reference_transcript.cons_vector).index(
+#                         next(x for x in sorted(reference_transcript.cons_vector) if x >= affected_cons_scores)) / len(
+#                 reference_transcript.cons_vector))
+#
+#             report = OncospliceAnnotator(reference_transcript, mutated_transcript, current_mutations[0])
+#             report['mut_id'] = mut_id
+#             report['splicing_engine'] = splicing_engine if splicing_engine is not None else 'None'
+#             # report['tis_engine'] = tis_engine if tis_engine is not None else 'None'
+#             # report['tis_pos'] = tis_candidate
+#             # report['tis_score'] = tis_score
+#             report['oncosplice_score'] = affected_cons_scores
+#             report['percentile'] = percentile
+#             report['isoform_id'] = short_hash_of_list(mutated_transcript.exons)
+#             report['isoform_prevalence'] = new_boundaries['path_weight']
+#             report['full_missplicing'] = missplicing.aberrant_splicing
+#             report['missplicing'] = missplicing.max_delta
+#             report['reference_resemblance'] = reference_gene_proteins.get(mutated_transcript.protein, None)
+#             results.append(report)
+#
+#     if len(results) == 0:
+#         # print("Nothing...")
+#         return pd.DataFrame()
+#
+#     return pd.DataFrame(results)[
+#         ['mut_id', 'transcript_id', 'isoform_id', 'primary_transcript', 'missplicing', 'full_missplicing',
+#          'exon_changes', 'splicing_codes', 'affected_exon', 'affected_intron', 'mutation_distance_from_5',
+#          'mutation_distance_from_3', 'missense_effect', 'codon_change', 'missense_position', 'reference_resemblance',
+#          'oncosplice_score', 'percentile', 'isoform_prevalence', 'reference_protein', 'variant_protein', 'splicing_engine']]
 
 
 def process_splicing_path(new_boundaries, reference_transcript, mutated_transcript,
@@ -451,11 +479,23 @@ def process_splicing_path(new_boundaries, reference_transcript, mutated_transcri
     """
     Processes a single alternative splicing path and returns an annotation report.
     """
+
+    base_results = {'mut_id': mut_id,
+                    'transcript_id': transcript_id,
+                    'execution_time': start_time,
+                    'status': '',
+                    'splicing_engine': splicing_engine if splicing_engine is not None else 'None',
+                    'isoform_id': short_hash_of_list(mutated_transcript.exons),
+                    }
     # Update acceptors and donors
     mutated_transcript.acceptors = new_boundaries['acceptors']
     mutated_transcript.donors = new_boundaries['donors']
     mutated_transcript.generate_mature_mrna().generate_protein()
 
+    if len(mutated_transcript.protein) <= 1:
+        base_results['status'] = 'Variant protein not viable.'
+        return base_results
+
     # Align reference and mutated proteins
     alignment = get_logical_alignment(reference_transcript.protein, mutated_transcript.protein)
     deleted, inserted = find_indels_with_mismatches_as_deletions(alignment.seqA, alignment.seqB)
@@ -490,7 +530,7 @@ def process_splicing_path(new_boundaries, reference_transcript, mutated_transcri
     return report
 
 
-def oncosplice_df(row, splicing_threshold=0.5, window_length=13,
+def oncosplice(row, splicing_threshold=0.5, window_length=13,
                   organism='hg38', splicing_engine='spliceai'):
     """
     Process a given mutation-transcript pair to analyze alternative splicing events
@@ -512,7 +552,11 @@ def oncosplice_df(row, splicing_threshold=0.5, window_length=13,
     """
 
     start_time = datetime.now().strftime("%Y-%m-%d %H:%M:%S")  # Log function start time
-    mut_id, transcript_id = row.mut_id, row.transcript_id
+    mut_id = row.mut_id
+    if 'transcript_id' not in row:
+        transcript_id = None
+    else:
+        transcript_id = row.transcript_id
 
     # Default response template (to ensure all IDs are included)
     base_result = {
@@ -528,7 +572,7 @@ def oncosplice_df(row, splicing_threshold=0.5, window_length=13,
         base_result['status'] = 'Gene not found'
         return pd.DataFrame([base_result])
 
-    mutations = [MutSeqMat.from_mutid(m) for m in mut_id.split('|')]
+    # mutations = [MutSeqMat.from_mutid(m) for m in mut_id.split('|')]
     reference_transcript = gene.transcript(transcript_id)
 
     if reference_transcript is None:
@@ -548,7 +592,7 @@ def oncosplice_df(row, splicing_threshold=0.5, window_length=13,
         return pd.DataFrame([base_result])
 
     # Filter mutations relevant to this transcript
-    current_mutations = [m for m in mutations if m in reference_transcript]
+    current_mutations = [m for m in mut_id.split('|') if m in reference_transcript]
     if not current_mutations:
         base_result['status'] = 'No relevant mutations'
         return pd.DataFrame([base_result])
@@ -602,7 +646,6 @@ def oncosplice_df(row, splicing_threshold=0.5, window_length=13,
     ]
 
 
-#
 # import asyncio
 # async def oncosplice_prototype(mut_id, splicing_threshold=0.5, protein_coding=True, primary_transcript=False,
 #                                window_length=13, organism='hg38', engine='spliceai', use_cons=True, require_cons=False):