Add ML predictor

docs/conf.py

@@ -70,7 +70,7 @@
 # The short X.Y version.
 version = '5.5'
 # The full version, including alpha/beta/rc tags.
-release = '5.5.2'
+release = '5.5.4'
 
 
 # The language for content autogenerated by Sphinx. Refer to documentation

docs/index.rst

@@ -63,9 +63,10 @@ New in Version |release|
 
 This is a bugfix release. It fixes the following problem(s):
 
-* Remove unused --top-score-metric2 parameter from pVACvector. by @ldhtnp in https://github.com/griffithlab/pVACtools/pull/1304
-* Speedup processing of AGFusion exon files. by @susannasiebert in https://github.com/griffithlab/pVACtools/pull/1301
-* Added biotype parameter to pVACvector when running with an input VCF. by @ldhtnp in https://github.com/griffithlab/pVACtools/pull/1297
+* Fix an issue where the output parsing of MHCflurryEL in pVACfuse and pVACsplice where only the smallest epitope length would have results. by @susannasiebert in https://github.com/griffithlab/pVACtools/pull/1320
+* Fix per-allele counts in the aggregated report. by @susannasiebert in https://github.com/griffithlab/pVACtools/pull/1316
+* Fix a few bugs where standalone commands weren't working with all_epitopes/filtered TSVs. by @susannasiebert in https://github.com/griffithlab/pVACtools/pull/1324
+* Update ``--top-score-metric2`` parameter help text for clarity and remove unnecessary usage. by @susannasiebert in https://github.com/griffithlab/pVACtools/pull/1321
 
 New in Version 5
 ----------------

docs/releases/5_5.rst

@@ -45,3 +45,21 @@ This is a bugfix release. It fixes the following problem(s):
 * Remove unused --top-score-metric2 parameter from pVACvector. by @ldhtnp in https://github.com/griffithlab/pVACtools/pull/1304
 * Speedup processing of AGFusion exon files. by @susannasiebert in https://github.com/griffithlab/pVACtools/pull/1301
 * Added biotype parameter to pVACvector when running with an input VCF. by @ldhtnp in https://github.com/griffithlab/pVACtools/pull/1297
+
+Version 5.5.3
+-------------
+
+This is a bugfix release. It fixes the following problem(s):
+
+* Fix a bug in the reference proteome similarity step that may result in an ``IndexError: string index out of range`` error. by @susannasiebert in https://github.com/griffithlab/pVACtools/pull/1313
+* Improve handling of inframe indels in repetitive region to upstream filter out variants that don't result in novel neoantigen candidates and to better find a matched wildtype epitope. by @susannasiebert in https://github.com/griffithlab/pVACtools/pull/1314
+
+Version 5.5.4
+-------------
+
+This is a bugfix release. It fixes the following problem(s):
+
+* Fix an issue where the output parsing of MHCflurryEL in pVACfuse and pVACsplice where only the smallest epitope length would have results. by @susannasiebert in https://github.com/griffithlab/pVACtools/pull/1320
+* Fix per-allele counts in the aggregated report. by @susannasiebert in https://github.com/griffithlab/pVACtools/pull/1316
+* Fix a few bugs where standalone commands weren't working with all_epitopes/filtered TSVs. by @susannasiebert in https://github.com/griffithlab/pVACtools/pull/1324
+* Update ``--top-score-metric2`` parameter help text for clarity and remove unnecessary usage. by @susannasiebert in https://github.com/griffithlab/pVACtools/pull/1321

pvactools/lib/binding_filter.py

@@ -7,15 +7,14 @@
 from pvactools.lib.run_utils import *
 
 class BindingFilter:
-    def __init__(self, input_file, output_file, binding_threshold, minimum_fold_change, top_score_metric, top_score_metric2, exclude_nas, allele_specific_cutoffs, percentile_threshold, percentile_threshold_strategy='conservative', file_type='pVACseq'):
+    def __init__(self, input_file, output_file, binding_threshold, minimum_fold_change, top_score_metric, exclude_nas, allele_specific_cutoffs, percentile_threshold, percentile_threshold_strategy='conservative', file_type='pVACseq'):
         self.input_file = input_file
         self.output_file = output_file
         self.binding_threshold = binding_threshold
         self.percentile_threshold = percentile_threshold
         self.percentile_threshold_strategy = percentile_threshold_strategy
         self.minimum_fold_change = minimum_fold_change
         self.top_score_metric = top_score_metric
-        self.top_score_metric2 = top_score_metric2
         self.exclude_nas = exclude_nas
         self.allele_specific_cutoffs = allele_specific_cutoffs
         self.file_type = file_type
@@ -105,12 +104,6 @@ def parser(cls, tool):
                  + "i.e. use the median MT ic50 binding score, fold change, and MT percentile of all chosen prediction methods).",
             default='median',
         )
-        parser.add_argument(
-            '-m2', '--top-score-metric2',
-            choices=['ic50','percentile'],
-            default='ic50',
-            help="Whether to use median/best IC50 or to use median/best percentile score."
-        )
         parser.add_argument(
             '--exclude-NAs',
             help="Exclude NA values from the filtered output.",

pvactools/lib/calculate_reference_proteome_similarity.py

@@ -216,14 +216,14 @@ def extract_n_mer(self, mt_peptide, wt_peptide):
         flanking_sequence_length = self.match_length - 1
         first_mut_aa_pos = 0
         for i in range(len(mt_peptide)):
-            if len(wt_peptide) < i:
+            if len(wt_peptide) <= i:
                 break
             if wt_peptide[i] != mt_peptide[i]:
                 first_mut_aa_pos = i
                 break
         last_mut_aa_pos = len(mt_peptide)
         for i in range(len(mt_peptide)):
-            if len(wt_peptide) < i:
+            if len(wt_peptide) <= i:
                 break
             if wt_peptide[i * -1] != mt_peptide[i * -1]:
                 last_mut_aa_pos = len(mt_peptide) - i
@@ -306,8 +306,8 @@ def _get_peptide(self, line, mt_records_dict, wt_records_dict):
                 epitope = line['Best Peptide']
                 subpeptide_position = int(line['Pos'])
             else:
-                identifier = line['Mutation']
-                epitope = line['MT Epitope Seq']
+                identifier = line['Index']
+                epitope = line['Epitope Seq']
                 subpeptide_position = int(line['Protein Position'])
             if identifier in mt_records_dict and identifier in wt_records_dict:
                 wt_peptide = wt_records_dict[identifier]

pvactools/lib/fasta_generator.py

@@ -283,9 +283,14 @@ def execute(self):
                 print("Warning. Mutant sequence contains unsupported amino acid U. Skipping entry {}".format(line['index']))
                 continue
 
-            if mutant_subsequence in wildtype_subsequence:
-                #This is not a novel peptide
-                continue
+            if self.epitope_length > 0:
+                neoepitopes = determine_neoepitopes(mutant_subsequence, self.epitope_length).values()
+                if all([neoepitope in wildtype_subsequence for neoepitope in neoepitopes]):
+                    #This variant does not result in any novel epitopes
+                    continue
+            else:
+                if mutant_subsequence in wildtype_subsequence:
+                    continue
 
             if len(wildtype_subsequence) < self.epitope_length or len(mutant_subsequence) < self.epitope_length:
                 continue