Add mix and match functionality for mutation calling (resolves #265) #283

davidstew · 2018-05-14T20:54:48Z

resolves #265

Added functionality so the user can select which mutation callers to execute. If no mutation callers are selected, the user must provide at least a mutation_vcf or a fusion_bedpe file.

arkal

Looking good. Minor edits needed. Also, Make sure you add an entry in MANUAL.md.

arkal · 2018-05-15T00:38:32Z

src/protect/mutation_calling/common.py

-    job.fileStore.logToMaster('Aggregated mutations for %s successfully' % univ_options['patient'])
-    return merged_snvs.rv()
+    chroms = {}
+    for caller in mutation_results:


Document why you've done this

arkal · 2018-05-15T00:39:25Z

src/protect/mutation_calling/common.py

@@ -105,25 +121,27 @@ def merge_perchrom_mutations(job, chrom, mutations, univ_options):
    # For now, let's just say 2 out of n need to call it.


remove these redundant comments

arkal · 2018-05-15T00:40:30Z

src/protect/mutation_calling/common.py

+        if 'strelka' + mut_type in perchrom_mutations:
+            perchrom_mutations['strelka'] = perchrom_mutations['strelka_' + mut_type]
+            perchrom_mutations.pop('strelka_' + mut_type)
+        if not perchrom_mutations:


This condition should occur before # process the strelka key

arkal · 2018-05-15T00:43:27Z

src/protect/mutation_calling/common.py

+            perchrom_mutations.pop('strelka_' + mut_type)
+        if not perchrom_mutations:
+            continue
+        majority = 1 if len(perchrom_mutations) <= 2 else len(perchrom_mutations) / 2


(len(perchrom_mutations) + 1) / 2

arkal · 2018-05-15T00:48:10Z

src/protect/pipeline/ProTECT.py

@@ -353,7 +353,7 @@ def _parse_config_file(job, config_file, max_cores=None):
    input_config = _add_default_entries(input_config, protect_defaults)

    # Flags to check for presence of encryption keys if required
-    gdc_inputs = ssec_encrypted = False
+    gdc_inputs = ssec_encrypted = sample_without_variants = False


# Flags to check for presence of encryption keys if required gdc_inputs = ssec_encrypted = False # Flag to check if a sample without an input vcf/bedpe was provided sample_without_variants = False

arkal · 2018-05-15T00:48:47Z

src/protect/pipeline/ProTECT.py

@@ -373,6 +373,8 @@ def _parse_config_file(job, config_file, max_cores=None):
                        raise ParameterError('Cannot run ProTECT using GDC RNA bams. Please fix '
                                             'sample %s' % sample_name)
                    gdc_inputs = True
+                if 'mutation_vcf' not in sample_set[sample_name] and 'fusion_bedpe' not in sample_set[sample_name]:


line too long

arkal

Can you implement the requested change, then squash and rebase?

After that, in a separate commit in the same PR, implement #286?

arkal · 2018-05-18T00:04:38Z

src/protect/mutation_calling/common.py

        # Process the strelka key
-        perchrom_mutations['strelka'] = perchrom_mutations['strelka_' + mut_type]
-        perchrom_mutations.pop('strelka_' + mut_type)
+        if 'strelka' + mut_type in perchrom_mutations:


This should be if 'strelka_' + mut_type in perchrom_mutations:

arkal · 2018-05-18T23:59:33Z

MANUAL.md

@@ -335,6 +335,11 @@ be substituted with S3 links. Descriptions for creating all files can be found i
            version: 1.2.0

    mutation_calling:
+        consensus:
+            indel_majority: None                                  -> Number of callers required to accept an indel.


remove the 'None'. Say default is to compute it dynamically. Same for snv

arkal · 2018-05-19T00:00:16Z

src/protect/pipeline/defaults.yaml

@@ -55,6 +55,9 @@ expression_estimation:
        version: 1.2.20

 mutation_calling:
+    consensus:
+        indel_majority: None


make this

consensus: indel_majority: snv_majority:

arkal · 2018-05-19T00:05:01Z

src/protect/pipeline/input_parameters.yaml

@@ -83,6 +83,9 @@ expression_estimation:
        # version: 1.2.0

 mutation_calling:
+    consensus:


Make this

consensus: # indel_majority: 1 # The number of callers required to accept an indel # snv_majority: 2 # The number of callers required to accept an snv

arkal · 2018-05-19T00:07:34Z

src/protect/pipeline/ProTECT.py

+            else:
+                count = 1 if tool_options['strelka']['run'] is True else 0
+            if tool_options['consensus'][mut_type + '_majority'] > count:
+                raise RuntimeError('Majority cannot be greater than the number of callers. Got number of %s callers '


This line is too long

resolves #265 Each mutation caller now has a ‘run’ option that can be set to False to skip using that caller.

resolves #286 Added snv_majority and indel_majority in a consensus tab under mutation_calling in input_parameters.yaml. If values are not provided, the majority will be dynamically figured out.

ghost assigned davidstew May 14, 2018

ghost added the in progress label May 14, 2018

arkal requested changes May 15, 2018

View reviewed changes

davidstew force-pushed the 265-mix-and-match-of-mutation-callers branch from f1e1f69 to 466c800 Compare May 15, 2018 23:23

arkal approved these changes May 16, 2018

View reviewed changes

arkal mentioned this pull request May 16, 2018

Allow user to specify majority #286

Open

arkal reviewed May 18, 2018

View reviewed changes

davidstew force-pushed the 265-mix-and-match-of-mutation-callers branch from 466c800 to f7266f6 Compare May 18, 2018 23:12

arkal requested changes May 19, 2018

View reviewed changes

arkal reviewed May 19, 2018

View reviewed changes

davidstew force-pushed the 265-mix-and-match-of-mutation-callers branch from f7266f6 to 85337d0 Compare May 19, 2018 00:28

davidstew added 2 commits June 5, 2018 16:22

Add mix and match functionality for mutation calling (resolves #265)

1945252

resolves #265 Each mutation caller now has a ‘run’ option that can be set to False to skip using that caller.

Make majority a user provided option (resolves #286)

b018a59

resolves #286 Added snv_majority and indel_majority in a consensus tab under mutation_calling in input_parameters.yaml. If values are not provided, the majority will be dynamically figured out.

davidstew force-pushed the 265-mix-and-match-of-mutation-callers branch from 85337d0 to b018a59 Compare June 5, 2018 23:23

Provide feedback

Saved searches

Use saved searches to filter your results more quickly

Add mix and match functionality for mutation calling (resolves #265) #283

Add mix and match functionality for mutation calling (resolves #265) #283

davidstew commented May 14, 2018

arkal left a comment

arkal May 15, 2018

arkal May 15, 2018

arkal May 15, 2018

arkal May 15, 2018

arkal May 15, 2018

arkal May 15, 2018

arkal left a comment

arkal May 18, 2018

arkal May 18, 2018

arkal May 19, 2018

arkal May 19, 2018

arkal May 19, 2018

		@@ -105,25 +121,27 @@ def merge_perchrom_mutations(job, chrom, mutations, univ_options):
		# For now, let's just say 2 out of n need to call it.

Add mix and match functionality for mutation calling (resolves #265) #283

Are you sure you want to change the base?

Add mix and match functionality for mutation calling (resolves #265) #283

Conversation

davidstew commented May 14, 2018

arkal left a comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

arkal left a comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment