Add parameters for penalty and threshold (#138)

* Add parameters for penalty and threshold

* Cleaning things up

* Cleaning up and changelog

CHANGELOG.md

@@ -11,6 +11,7 @@ Try to use the following format:
 ## [unreleased]
 - Fixed wrong models when chromosome X was named `chrX` and not `X` 
 - Added GitHub Actions workflows for automatic publishing to PyPI on release, and keep a changelog reminder ([#136](https://github.com/Clinical-Genomics/genmod/pull/136))
+- Optional user defined threshold and penalty for compound scoring
 
 ## [3.8.3]
 

genmod/commands/score_compounds.py

@@ -45,8 +45,10 @@
                     is_flag=True,
                     help='If variants are annotated with the Variant Effect Predictor.'
 )
+@click.option('--threshold', type=int, help="Threshold for model-dependent penalty if no compounds with passing score", default=9)
+@click.option('--penalty', type=int, help="Penalty applied together with --threshold", default=6)
 @click.pass_context
-def compound(context, variant_file, silent, outfile, vep, processes, temp_dir):
+def compound(context, variant_file, silent, outfile, vep, threshold: int, penalty: int, processes, temp_dir):
     """
     Score compound variants in a vcf file based on their rank score.
     """
@@ -110,6 +112,8 @@ def compound(context, variant_file, silent, outfile, vep, processes, temp_dir):
             task_queue=variant_queue,
             results_queue=results,
             individuals=individuals,
+            threshold=threshold,
+            penalty=penalty,
         )
         for i in range(num_scorers)
     ]

genmod/score_variants/compound_scorer.py

@@ -90,7 +90,7 @@ class CompoundScorer(Process):
     the results queue.
     """
 
-    def __init__(self, task_queue, results_queue, individuals):
+    def __init__(self, task_queue, results_queue, individuals, threshold: int, penalty: int):
         """
         Initialize the VariantAnnotator
         
@@ -119,6 +119,9 @@ def __init__(self, task_queue, results_queue, individuals):
         logger.debug("Setting up individuals")
         self.individuals = individuals
 
+        self.threshold = threshold
+        self.penalty = penalty
+
         if len(self.individuals) == 1:
             self.models = ['AR_comp', 'AR_comp_dn', 'AD', 'AD_dn']
         else:
@@ -235,7 +238,7 @@ def run(self):
                         for compound_id in compound_list:
                             compound_rank_score = rank_scores[rank_score_type][compound_id]
                             if compound_rank_score > get_rank_score(rank_score_type=rank_score_type,
-                                                                    threshold=9,
+                                                                    threshold=self.threshold,
                                                                     min_rank_score_value=variant_rankscore_normalization_bounds[variant_id][0],
                                                                     max_rank_score_value=variant_rankscore_normalization_bounds[variant_id][1]
                                                                     ):
@@ -246,7 +249,7 @@ def run(self):
                             logger.debug("correcting rank score for {0}".format(
                                 variant_id))
                             current_rank_score -= get_rank_score_as_magnitude(rank_score_type=rank_score_type,
-                                                                              rank_score=6,
+                                                                              rank_score=self.penalty,
                                                                               min_rank_score_value=variant_rankscore_normalization_bounds[variant_id][0],
                                                                               max_rank_score_value=variant_rankscore_normalization_bounds[variant_id][1]
                                                                               )