def audit_biosample_depleted_term_match(value, system):
'''
The depleted_in_term_name and depleted_in_term_name
should be concordant. This should be a calcualted field.
If one exists, the other should. This should be handled in the schema.
'''
if value['status'] == 'deleted':
return
if 'depleted_in_term_name' not in value:
return
if len(value['depleted_in_term_name']) != len(
value['depleted_in_term_id']):
detail = 'Biosample {} has a depleted_in_term_name array and depleted_in_term_id array of differing lengths'.format(
value['@id'])
raise AuditFailure('mismatched depleted_in_term length',
detail,
level='ERROR')
return
for i, dep_term in enumerate(value['depleted_in_term_name']):
if (term_mapping[dep_term]) != (value['depleted_in_term_id'][i]):
detail = 'Biosample {} has a mismatch between {} and {}'.format(
value['@id'], dep_term, value['depleted_in_term_id'][i])
raise AuditFailure('mismatched depleted_in_term',
detail,
level='ERROR')
def audit_experiment_replicated(value, system):
'''
Experiments in ready for review or release ready state should be replicated. If not,
wranglers should check with lab as to why before release.
'''
if value['status'] not in [
'released', 'release ready', 'ready for review'
]:
return
'''
Excluding single cell isolation experiments from the replication requirement
'''
if value['assay_term_name'] == 'single cell isolation followed by RNA-seq':
return
num_bio_reps = set()
for rep in value['replicates']:
num_bio_reps.add(rep['biological_replicate_number'])
if len(num_bio_reps) <= 1:
if value['status'] in ['released']:
detail = 'Experiment {} has only one biological replicate and is released. Check for proper annotation of this state in the metadata'.format(
value['@id'])
raise AuditFailure('unreplicated experiment',
detail,
level='DCC_ACTION')
if value['status'] in ['ready for review', 'release ready']:
detail = 'Experiment {} has only one biological replicate, more than one is typically expected before release'.format(
value['@id'])
raise AuditFailure('unreplicated experiment',
detail,
level='WARNING')
def audit_file_paired_ended_run_type(value, system):
'''
Audit to catch those files that were upgraded to have run_type = paired ended
resulting from its migration out of replicate but lack the paired_end property
to specify which read it is. This audit will also catch the case where run_type
= paired-ended but there is no paired_end = 2 due to registeration error.
'''
if value['status'] in ['deleted', 'replaced', 'revoked', 'upload failed']:
return
if value['file_format'] not in ['fastq', 'fasta', 'csfasta']:
return
if (value['output_type'] == 'reads') and (value.get('run_type')
== 'paired-ended'):
if 'paired_end' not in value:
detail = 'File {} has a paired-ended run_type but is missing its paired_end value'.format(
value['@id'])
raise AuditFailure('missing paired_end',
detail,
level='DCC_ACTION')
if (value['paired_end'] == 1) and 'paired_with' not in value:
detail = 'File {} has a paired-ended run_type but is missing a paired_end=2 mate'.format(
value['@id'])
raise AuditFailure('missing mate pair', detail, level='DCC_ACTION')
def audit_experiment_control(value, system):
'''
Certain assay types (ChIP-seq, ...) require possible controls with a matching biosample.
Of course, controls do not require controls.
'''
if value['status'] in ['deleted', 'proposed']:
return
# Currently controls are only be required for ChIP-seq
if value.get('assay_term_name') not in controlRequiredAssayList:
return
# We do not want controls
if 'target' in value and 'control' in value['target']['investigated_as']:
return
if value['possible_controls'] == []:
detail = '{} experiments require a value in possible_control'.format(
value['assay_term_name'])
raise AuditFailure('missing possible_controls',
detail,
level='NOT_COMPLIANT')
for control in value['possible_controls']:
if control.get('biosample_term_id') != value.get('biosample_term_id'):
detail = 'Control {} is for {} but experiment is done on {}'.format(
control['@id'], control.get('biosample_term_name'),
value['biosample_term_name'])
raise AuditFailure('mismatched control', detail, level='ERROR')
def audit_experiment_replicates_biosample(value, system):
if value['status'] in ['deleted', 'replaced', 'revoked']:
return
biological_replicates_dict = {}
biosamples_list = []
for rep in value['replicates']:
bio_rep_num = rep['biological_replicate_number']
tech_rep_num = rep['technical_replicate_number']
if 'library' in rep and 'biosample' in rep['library']:
biosample = rep['library']['biosample']
if not bio_rep_num in biological_replicates_dict:
biological_replicates_dict[bio_rep_num] = biosample[
'accession']
if biosample['accession'] in biosamples_list:
detail = 'Experiment {} has multiple biological replicates associated with the same biosample {}'.format(
value['@id'], biosample['@id'])
raise AuditFailure(
'biological replicates with identical biosample',
detail,
level='DCC_ACTION')
else:
biosamples_list.append(biosample['accession'])
else:
if biosample['accession'] != biological_replicates_dict[
bio_rep_num]:
detail = 'Experiment {} has technical replicates associated with the different biosamples'.format(
value['@id'])
raise AuditFailure(
'technical replicates with not identical biosample',
detail,
level='DCC_ACTION')
def audit_library_depleted_in(value, system):
'''
If there is a depleted_term_name or term_id,
both should exist - should be handled by schema
They should match each other.
This should also be replaced by a calculated field
'''
if value['status'] in ['deleted']:
return
if not value['depleted_in_term_name'] or not value['depleted_in_term_id']:
return
if len(value['depleted_in_term_name']) != len(value['depleted_in_term_id']):
detail = 'Library {} has depleted_in_term_name array and depleted_in_term_id array of differing lengths'.format(
value['@id'])
yield AuditFailure('depleted_in length mismatch', detail, level='ERROR')
for i, dep_term in enumerate(value['depleted_in_term_id']):
if dep_term == value['nucleic_acid_term_id']:
detail = 'Library {} of type {} cannot be depleted in {}'.format(
value['@id'],
value['nucleic_acid_term_id'],
value['depleted_in_term_id'][i])
yield AuditFailure('invalid depleted_in_term_id', detail, level='ERROR')
expected = moleculeDict[value['depleted_in_term_name'][i]]
if expected != value['depleted_in_term_id'][i]:
detail = 'Library {} has mismatch between {} - {}'.format(
value['@id'],
value['depleted_in_term_name'][i],
value['depleted_in_term_id'][i])
yield AuditFailure('mismatched depleted_in_term', detail, level='ERROR')
def audit_file_controlled_by(value, system):
'''
A fastq in a ChIP-seq experiment should have a controlled_by
'''
if value['status'] in ['deleted', 'replaced', 'revoked']:
return
if value['dataset'].get('assay_term_name') not in [
'ChIP-seq', 'RAMPAGE', 'CAGE',
'shRNA knockdown followed by RNA-seq'
]:
return
if 'target' in value['dataset'] and 'control' in value['dataset'][
'target'].get('investigated_as', []):
return
if 'controlled_by' not in value:
value['controlled_by'] = []
if (value['controlled_by'] == []) and (value['file_format'] in ['fastq']):
detail = 'Fastq file {} from {} requires controlled_by'.format(
value['@id'], value['dataset']['assay_term_name'])
raise AuditFailure('missing controlled_by',
detail,
level='NOT_COMPLIANT')
possible_controls = value['dataset'].get('possible_controls')
biosample = value['dataset'].get('biosample_term_id')
for ff in value['controlled_by']:
control_bs = ff['dataset'].get('biosample_term_id')
if control_bs != biosample:
detail = 'File {} has a controlled_by file {} with conflicting biosample {}'.format(
value['@id'], ff['@id'], control_bs)
raise AuditFailure('mismatched controlled_by',
detail,
level='ERROR')
return
if ff['file_format'] != value['file_format']:
detail = 'File {} with file_format {} has a controlled_by file {} with file_format {}'.format(
value['@id'], value['file_format'], ff['@id'],
ff['file_format'])
raise AuditFailure('mismatched controlled_by',
detail,
level='ERROR')
if (possible_controls is None) or (ff['dataset']['@id']
not in possible_controls):
detail = 'File {} has a controlled_by file {} with a dataset {} that is not in possible_controls'.format(
value['@id'], ff['@id'], ff['dataset']['@id'])
raise AuditFailure('mismatched controlled_by',
detail,
level='DCC_ACTION')
def audit_experiment_assay(value, system):
'''
Experiments should have assays with valid ontologies term ids and names that
are a valid synonym.
'''
if value['status'] == 'deleted':
return
if 'assay_term_id' not in value:
detail = 'Experiment {} is missing assay_term_id'.format(value['@id'])
yield AuditFailure('missing assay information', detail, level='ERROR')
return
# This should be a dependancy
if 'assay_term_name' not in value:
detail = 'Experiment {} is missing assay_term_name'.format(
value['@id'])
yield AuditFailure('missing assay information', detail, level='ERROR')
return
# This should be a dependancy
ontology = system['registry']['ontology']
term_id = value.get('assay_term_id')
term_name = value.get('assay_term_name')
if term_id.startswith('NTR:'):
detail = 'Assay_term_id is a New Term Request ({} - {})'.format(
term_id, term_name)
yield AuditFailure('NTR assay', detail, level='DCC_ACTION')
return
if term_id not in ontology:
detail = 'Assay_term_id {} is not found in cached version of ontology'.format(
term_id)
yield AuditFailure('assay_term_id not in ontology',
term_id,
level='DCC_ACTION')
return
ontology_term_name = ontology[term_id]['name']
modifed_term_name = term_name + ' assay'
if (ontology_term_name != term_name and term_name not in ontology[term_id]['synonyms']) and \
(ontology_term_name != modifed_term_name and
modifed_term_name not in ontology[term_id]['synonyms']):
detail = 'Experiment has a mismatch between assay_term_name "{}" and assay_term_id "{}"'.format(
term_name,
term_id,
)
yield AuditFailure('mismatched assay_term_name',
detail,
level='DCC_ACTION')
return
def audit_biosample_concordance(value, system):
'''
The biosample details of the experiment of a replicate and the library.biosample of a replicate
need to match.
'''
if value.get('status') in ['deleted', 'replaced']:
return
if 'library' not in value:
return
if 'biosample' not in value['library']:
return
exp = value['experiment']['@id']
exp_type = value['experiment'].get('biosample_type')
exp_name = value['experiment'].get('biosample_term_name')
exp_id = value['experiment'].get('biosample_term_id')
bio = value['library']['biosample']['@id']
bs_type = value['library']['biosample'].get('biosample_type')
bs_name = value['library']['biosample'].get('biosample_term_name')
bs_id = value['library']['biosample'].get('biosample_term_id')
if bs_type != exp_type:
detail = '{} has mismatched biosample_type: {}, but {} in {}'.format(
exp,
exp_type,
bs_type,
bio
)
yield AuditFailure('mismatched biosample_type', detail, level='ERROR')
if bs_name != exp_name:
detail = '{} has mismatched biosample_term_name: {}, but {} in {}'.format(
exp,
exp_name,
bs_name,
bio
)
yield AuditFailure('mismatched biosample_term_name', detail, level='ERROR')
if bs_id != exp_id:
detail = '{} has mismatched biosample_term_id: {}, but {} in {}'.format(
bio,
exp_id,
bs_id,
bio
)
yield AuditFailure('mismatched biosample_term_id', detail, level='ERROR')
def audit_antibody_characterization_unique_reviews(value, system):
'''
Make sure primary characterizations have unique lane, biosample_term_id and
organism combinations for characterization reviews
'''
if (value['status'] in [
'deleted', 'not submitted for review by lab', 'in progress',
'not reviewed'
]):
return
if 'secondary_characterization_method' in value:
return
unique_reviews = set()
for review in value['characterization_reviews']:
lane = review['lane']
term_id = review['biosample_term_id']
organism = review['organism']
review_lane = frozenset([lane, term_id, organism])
if review_lane not in unique_reviews:
unique_reviews.add(review_lane)
else:
detail = 'Characterization_review.lane {} is a duplicate review for {} - {}'.format(
lane, term_id, organism)
raise AuditFailure('duplicate lane review', detail, level='ERROR')
def audit_experiment_spikeins(value, system):
'''
All ENCODE 3 long (>200) RNA-seq experiments should specify their spikeins.
The spikeins specified should have datasets of type spikeins.
The spikeins datasets should have a fasta file, a document, and maybe a tsv
'''
if value['status'] in ['deleted', 'replaced']:
return
if value.get('assay_term_name') != 'RNA-seq':
return
for rep in value['replicates']:
lib = rep.get('library')
if lib is None:
continue
size_range = lib.get('size_range')
if size_range != '>200':
continue
spikes = lib.get('spikeins_used')
if (spikes is None) or (spikes == []):
detail = 'Library {} is in an RNA-seq experiment and has size_range >200. It requires a value for spikeins_used'.format(
lib['@id'])
yield AuditFailure('missing spikeins_used',
detail,
level='NOT_COMPLIANT')
def audit_file_format_specifications(value, system):
for doc in value.get('file_format_specifications', []):
if doc['document_type'] != "file format specification":
detail = 'File {} has document {} not of type file format specification'.format(
value['@id'], doc['@id'])
raise AuditFailure('wrong document_type', detail, level='ERROR')
def audit_biosample_donor(value, system):
'''
A biosample should have a donor.
The organism of donor and biosample should match.
Pooled_from biosamples do not need donors??
'''
if value['status'] in ['deleted']:
return
if ('donor' not in value) and (value['pooled_from']):
return
if ('donor' not in value) and (not value['pooled_from']):
detail = 'Biosample {} requires a donor'.format(value['@id'])
raise AuditFailure('missing donor', detail, level='ERROR')
return
donor = value['donor']
if value['organism']['name'] != donor['organism']['name']:
detail = 'Biosample {} is organism {}, yet its donor {} is organism {}. Biosamples require a donor of the same species'.format(
value['@id'], value['organism']['name'], donor['@id'],
donor['organism']['name'])
raise AuditFailure('mismatched organism', detail, level='ERROR')
if 'mutated_gene' not in donor:
return
if value['organism']['name'] != donor['mutated_gene']['organism']['name']:
detail = 'Biosample {} is organism {}, but its donor {} mutated_gene is in {}. Donor mutated_gene should be of the same species as the donor and biosample'.format(
value['@id'], value['organism']['name'], donor['@id'],
donor['mutated_gene']['organism']['name'])
raise AuditFailure('mismatched mutated_gene organism',
detail,
level='ERROR')
for i in donor['mutated_gene']['investigated_as']:
if i in [
'histone modification', 'tag', 'control',
'recombinant protein', 'nucleotide modification',
'other post-translational modification'
]:
detail = 'Donor {} has an invalid mutated_gene {}. Donor mutated_genes should not be tags, controls, recombinant proteins or modifications'.format(
donor['@id'], donor['mutated_gene']['name'])
raise AuditFailure('invalid donor mutated_gene',
detail,
level='ERROR')
def audit_biosample_transfection_type(value, system):
'''
A biosample with constructs or rnais should have a
transfection_type
'''
if value['status'] == 'deleted':
return
if (value['rnais']) and ('transfection_type' not in value):
detail = 'Biosample {} with a value for RNAi requires transfection_type'.format(
value['@id'])
raise AuditFailure('missing transfection_type', detail, level='ERROR')
if (value['constructs']) and ('transfection_type' not in value):
detail = 'Biosample {} with a value for construct requires transfection_type'.format(
value['@id'])
raise AuditFailure('missing transfection_type', detail, level='ERROR')
def audit_file_size(value, system):
if value['status'] in ['deleted', 'replaced', 'uploading', 'revoked']:
return
if 'file_size' not in value:
detail = 'File {} requires a value for file_size'.format(value['@id'])
raise AuditFailure('missing file_size', detail, level='DCC_ACTION')
def audit_antibody_characterization_status(value, system):
'''
Make sure the lane_status matches
the characterization status
'''
if 'secondary_characterization_method' in value:
return
if (value['status'] in [
"deleted", "not submitted for review by lab", 'in progress',
'not reviewed'
]):
if 'characterization_reviews' in value:
'''If any of these statuses, we shouldn't have characterization_reviews'''
detail = 'Antibody_characterization.status of {} is incompatible with having a value for characterization_reviews'.format(
value['status'])
raise AuditFailure('unexpected characterization_reviews',
detail,
level='WARNING')
else:
return
'''Check each of the lane_statuses in characterization_reviews for an appropriate match'''
has_compliant_lane = False
is_pending = False
if value['status'] == 'pending dcc review':
is_pending = True
for lane in value['characterization_reviews']:
if (is_pending and lane['lane_status'] != 'pending dcc review') or (
not is_pending
and lane['lane_status'] == 'pending dcc review'):
detail = 'A lane.status of {} is incompatible with antibody_characterization.status of pending dcc review'.format(
lane['lane_status'])
raise AuditFailure('mismatched lane status',
detail,
level='WARNING')
continue
if lane['lane_status'] == 'compliant':
has_compliant_lane = True
if has_compliant_lane and value['status'] != 'compliant':
detail = 'A lane.status of {} is incompatible with antibody_characterization status of {}'.format(
lane['lane_status'], value['status'])
raise AuditFailure('mismatched lane status',
detail,
level='DCC_ACTION')
def audit_experiment_release_date(value, system):
'''
Released experiments need release date.
This should eventually go to schema
'''
if value['status'] == 'released' and 'date_released' not in value:
detail = 'Experiment {} is released and requires a value in date_released'.format(
value['@id'])
raise AuditFailure('missing date_released', detail, level='DCC_ACTION')
def audit_experiment_ChIP_control(value, system):
if value['status'] in [
'deleted', 'proposed', 'preliminary', 'replaced', 'revoked'
]:
return
# Currently controls are only be required for ChIP-seq
if value.get('assay_term_name') != 'ChIP-seq':
return
# We do not want controls
if 'target' in value and 'control' in value['target']['investigated_as']:
return
if not value['possible_controls']:
return
num_IgG_controls = 0
for control in value['possible_controls']:
if ('target' not in control) or (
'control' not in control['target']['investigated_as']):
detail = 'Experiment {} is ChIP-seq but its control {} is not linked to a target with investigated.as = control'.format(
value['@id'], control['@id'])
raise AuditFailure('invalid possible_control',
detail,
level='ERROR')
if not control['replicates']:
continue
if 'antibody' in control['replicates'][0]:
num_IgG_controls += 1
# If all of the possible_control experiments are mock IP control experiments
if num_IgG_controls == len(value['possible_controls']):
if value.get('assay_term_name') == 'ChIP-seq':
# The binding group agreed that ChIP-seqs all should have an input control.
detail = 'Experiment {} is ChIP-seq and requires at least one input control, as agreed upon by the binding group. {} is not an input control'.format(
value['@id'], control['@id'])
raise AuditFailure('missing input control',
detail,
level='NOT_COMPLIANT')
def audit_library_biosample(value, system):
'''
The library should be linked to biosample
'''
if value['status'] in ['deleted']:
return
if 'biosample' not in value:
detail = 'Library {} has no biosample'.format(
value['@id'])
raise AuditFailure('missing biosample', detail, level='ERROR')
def audit_paired_with(value, system):
'''
A file with a paired_end needs a paired_with.
Should be handled in the schema.
A paired_with should be the same replicate
'''
if value['status'] in ['deleted', 'replaced', 'revoked']:
return
if 'paired_end' not in value:
return
if 'paired_with' not in value:
detail = 'File {} has paired_end = {}. It requires a paired file'.format(
value['@id'], value['paired_end'])
raise AuditFailure('missing paired_with',
detail,
level='NOT_COMPLIANT')
if 'replicate' not in value['paired_with']:
return
if 'replicate' not in value:
detail = 'File {} has paired_end = {}. It requires a replicate'.format(
value['@id'], value['paired_end'])
raise AuditFailure('missing replicate', detail, level='DCC_ACTION')
if value['replicate'] != value['paired_with']['replicate']:
detail = 'File {} has replicate {}. It is paired_with file {} with replicate {}'.format(
value['@id'], value.get('replicate'), value['paired_with']['@id'],
value['paired_with'].get('replicate'))
raise AuditFailure('mismatched paired_with', detail, level='ERROR')
if value['paired_end'] == '1':
context = system['context']
paired_with = context.get_rev_links('paired_with')
if len(paired_with) > 1:
detail = 'Paired end 1 file {} paired_with by multiple paired end 2 files: {!r}'.format(
value['@id'],
paired_with,
)
raise AuditFailure('multiple paired_with', detail, level='ERROR')
def audit_biosample_term(value, system):
'''
Biosample_term_id and biosample_term_name
and biosample_type should all be present.
This should be handled by schemas.
Biosample_term_id should be in the ontology.
Biosample_term_name should match biosample_term_id.
'''
if value['status'] in ['deleted']:
return
if 'biosample_term_id' not in value:
return
ontology = system['registry']['ontology']
term_id = value['biosample_term_id']
term_name = value.get('biosample_term_name')
if term_id.startswith('NTR:'):
detail = 'Biosample {} has a New Term Request {} - {}'.format(
value['@id'], term_id, term_name)
raise AuditFailure('NTR biosample', detail, level='DCC_ACTION')
if term_id not in ontology:
detail = 'Biosample {} has biosample_term_id of {} which is not in ontology'.format(
value['@id'], term_id)
raise AuditFailure('term_id not in ontology',
term_id,
level='DCC_ACTION')
ontology_term_name = ontology[term_id]['name']
if ontology_term_name != term_name and term_name not in ontology[term_id][
'synonyms']:
detail = 'Biosample {} has a mismatch between biosample_term_id "{}" and biosample_term_name "{}"'.format(
value['@id'],
term_id,
term_name,
)
raise AuditFailure('mismatched biosample_term',
detail,
level='DCC_ACTION')
def audit_references_for_publication(value, system):
'''
For datasets of type publication, there should be references. Those that
do not should be earmarked so they can be added once the publication
has been accepted
'''
if value['status'] in ['deleted', 'replaced', 'revoked', 'preliminary']:
return
if (value['dataset_type'] == 'publication') and (not value['references']):
detail = 'publication dataset missing a reference to a publication'
raise AuditFailure('missing reference', detail, level='WARNING')
def audit_antibody_characterization_review(value, system):
'''
Make sure that biosample terms are in ontology
for each characterization_review.
'''
if (value['status'] in [
'not reviewed', 'not submitted for review by lab', 'deleted',
'in progress'
]):
return
if 'secondary_characterization_method' in value:
return
if value['characterization_reviews']:
ontology = system['registry']['ontology']
for review in value['characterization_reviews']:
term_id = review['biosample_term_id']
term_name = review['biosample_term_name']
if term_id.startswith('NTR:'):
detail = 'Antibody_characterization {} contains a New Term Request {} - {}'.format(
value['@id'], term_id, term_name)
raise AuditFailure('NTR biosample', detail, level='DCC_ACTION')
if term_id not in ontology:
detail = 'Antibody characterization {} contains a biosample_term_id {} that is not in the ontology'.format(
value['@id'], term_id)
raise AuditFailure('term_id not in ontology',
term_id,
level='DCC_ACTION')
ontology_term_name = ontology[term_id]['name']
if ontology_term_name != term_name and term_name not in ontology[
term_id]['synonyms']:
detail = 'Antibody characterization {} has a mismatched term {} - {} expected {}'.format(
value['@id'], term_id, term_name, ontology_term_name)
raise AuditFailure('mismatched term_name',
detail,
level='ERROR')
def audit_analysis_steps_closure(value, system):
''' The analysis_steps list should include all of a steps ancestors.
'''
ids = {step['@id'] for step in value['analysis_steps']}
parents = {
parent
for step in value['analysis_steps']
for parent in step.get('parents', [])
}
diff = parents.difference(ids)
if diff:
detail = ', '.join(sorted(diff))
raise AuditFailure('incomplete analysis_steps', detail, level='ERROR')
def audit_experiment_replicates_with_no_libraries(value, system):
if value['status'] in ['deleted', 'replaced', 'revoked']:
return
if len(value['replicates']) == 0:
return
for rep in value['replicates']:
if 'library' not in rep:
detail = 'Experiment {} has a replicate {}, that has no library associated with'.format(
value['@id'], rep['@id'])
yield AuditFailure('replicate with no library',
detail,
level='DCC_ACTION')
return
def audit_library_RNA_size_range(value, system):
'''
An RNA library should have a size_range specified.
This needs to accomodate the rfa
'''
if value['status'] in ['deleted']:
return
RNAs = ['SO:0000356', 'SO:0000871']
if (value['nucleic_acid_term_id'] in RNAs) and ('size_range' not in value):
detail = 'RNA library {} requires a value for size_range'.format(value['@id'])
raise AuditFailure('missing size_range', detail, level='ERROR')
def audit_file_read_length(value, system):
'''
Reads files should have a read_length
'''
if value['status'] in ['deleted', 'replaced', 'revoked']:
return
if value['output_type'] != 'reads':
return
if 'read_length' not in value:
detail = 'Reads file {} missing read_length'.format(value['@id'])
raise AuditFailure('missing read_length', detail, level='ERROR')
def audit_antibody_characterization_target(value, system):
'''
Make sure that target in characterization
matches target of antibody
'''
antibody = value['characterizes']
target = value['target']
if 'recombinant protein' in target['investigated_as']:
prefix = target['label'].split('-')[0]
unique_antibody_target = set()
unique_investigated_as = set()
for antibody_target in antibody['targets']:
label = antibody_target['label']
unique_antibody_target.add(label)
for investigated_as in antibody_target['investigated_as']:
unique_investigated_as.add(investigated_as)
if 'tag' not in unique_investigated_as:
detail = 'Antibody {} is not for a tagged protein, yet target is investigated_as a recombinant protein'.format(
antibody['@id'])
raise AuditFailure('not tagged antibody', detail, level='ERROR')
else:
if prefix not in unique_antibody_target:
detail = '{} is not found in target list for antibody {}'.format(
prefix, antibody['@id'])
raise AuditFailure('mismatched tag target',
detail,
level='ERROR')
else:
target_matches = False
for antibody_target in antibody['targets']:
if target['name'] == antibody_target.get('name'):
target_matches = True
if not target_matches:
detail = 'Target {} is not found in target list for antibody {}'.format(
target['name'], antibody['@id'])
raise AuditFailure('mismatched target', detail, level='ERROR')
def audit_run_type(value, system):
'''
A fastq file or a fasta file need to specify run_type.
This was attempted to be a dependancy and didn't happen.
'''
if value['status'] in ['deleted', 'replaced', 'revoked']:
return
if value['file_format'] not in ['fastq', 'fasta']:
return
if 'run_type' not in value:
detail = 'File {} has file_format {}. It requires a value for run_type'.format(
value['@id'], value['file_format'])
raise AuditFailure('missing run_type', detail, level='NOT_COMPLIANT')
def audit_file_platform(value, system):
'''
A raw data file should have a platform specified.
Should be in the schema.
'''
if value['status'] in ['deleted', 'replaced']:
return
if value['file_format'] not in raw_data_formats:
return
if 'platform' not in value:
detail = 'Raw data file {} missing platform information'.format(
value['@id'])
raise AuditFailure('missing platform', detail, level='ERROR')
