issue #76 seqauto API step 1 - serializers/initial api

genes/serializers.py

@@ -1,9 +1,10 @@
 from rest_framework import serializers
 
 from genes.models import GeneInfo, GeneListCategory, GeneList, Gene, Transcript, GeneListGeneSymbol, \
-    GeneAnnotationRelease, SampleGeneList, ActiveSampleGeneList, GeneSymbol, TranscriptVersion, GeneVersion, HGNC
+    GeneAnnotationRelease, SampleGeneList, ActiveSampleGeneList, GeneSymbol, TranscriptVersion, GeneVersion, HGNC, \
+    GeneCoverageCollection, GeneCoverageCanonicalTranscript
 from snpdb.models import Company
-from snpdb.serializers import UserSerializer, GenomeBuildSerializer
+from snpdb.serializers import GenomeBuildSerializer
 
 
 class GeneSymbolSerializer(serializers.ModelSerializer):
@@ -74,7 +75,7 @@ class Meta:
 
 class GeneListSerializer(serializers.ModelSerializer):
     category = GeneListCategorySerializer()
-    user = UserSerializer()
+    user = serializers.StringRelatedField()
     genelistgenesymbol_set = GeneListGeneSymbolSerializer(many=True)
     can_write = serializers.SerializerMethodField()
     absolute_url = serializers.URLField(source='get_absolute_url', read_only=True)
@@ -131,14 +132,34 @@ class Meta:
 
 
 class SampleGeneListSerializer(serializers.ModelSerializer):
-    active = serializers.SerializerMethodField()
+    active = serializers.SerializerMethodField(read_only=True)
+    gene_list = GeneListSerializer()
 
     class Meta:
         model = SampleGeneList
-        fields = ('pk', 'visible', 'active')
+        fields = ('pk', 'visible', 'gene_list', 'active')
 
     def get_active(self, obj):
         try:
             return obj.sample.activesamplegenelist.sample_gene_list == obj
         except ActiveSampleGeneList.DoesNotExist:
             return False
+
+
+class GeneCoverageCanonicalTranscriptSerializer(serializers.ModelSerializer):
+    transcript_version = TranscriptVersionSerializer()
+
+    class Meta:
+        model = GeneCoverageCanonicalTranscript
+        exclude = ("gene_coverage_collection", )
+        #fields = "__all__"
+
+
+class GeneCoverageCollectionSerializer(serializers.ModelSerializer):
+    genome_build = GenomeBuildSerializer()
+    genecoveragecanonicaltranscript_set = GeneCoverageCanonicalTranscriptSerializer(many=True)
+
+    class Meta:
+        model = GeneCoverageCollection
+        # TODO: Check if "__all__" also does related automatically?
+        fields = ["path", "data_state", "genome_build", "genecoveragecanonicaltranscript_set"]

seqauto/models/models_seqauto.py

@@ -389,7 +389,10 @@ def __str__(self):
 
 
 class SequencingSample(models.Model):
-    """ Represents a row in a SampleSheet.csv """
+    """ Represents a row in a SampleSheet.csv
+
+        As it's not a file, not a SeqAutoRecord
+     """
     sample_sheet = models.ForeignKey(SampleSheet, on_delete=CASCADE)
     sample_id = models.TextField()
     # sample_name is used to name files. In MiSeq/NextSeq samplesheet you can add names.
@@ -973,7 +976,12 @@ def __str__(self):
 
 
 class QCGeneList(SeqAutoRecord):
-    """ This represents a text file containing genes which will be used for initial pass and QC filters """
+    """ This represents a text file containing genes which will be used for initial pass and QC filters
+
+        The reason we have both a sample_gene_list and a custom_text_gene_list is because we wanted to
+        represent the text from a file on disk. I think we probably could have gotten around that as
+        SeqAutoRecord contains a hash - could maybe remove that and just use gene list
+    """
     qc = models.ForeignKey(QC, on_delete=CASCADE)
     custom_text_gene_list = models.OneToOneField(CustomTextGeneList, null=True, on_delete=SET_NULL)
     sample_gene_list = models.ForeignKey(SampleGeneList, null=True, on_delete=SET_NULL)

seqauto/serializers/__init__.py

@@ -0,0 +1,3 @@
+from .enrichment_kit_serializers import *
+from .seqauto_serializers import *
+from .seqauto_serializers import *

seqauto/serializers/enrichment_kit_serializers.py

@@ -0,0 +1,39 @@
+from rest_framework import serializers
+
+from genes.serializers import GeneListSerializer
+from seqauto.models import EnrichmentKit
+
+
+
+class EnrichmentKitSerializer(serializers.ModelSerializer):
+    gene_list = GeneListSerializer()
+    enrichment_kit_type = serializers.SerializerMethodField()
+    manufacturer = serializers.StringRelatedField()
+    __str__ = serializers.SerializerMethodField(read_only=True)
+
+    class Meta:
+        model = EnrichmentKit
+        fields = ('pk', 'name', 'version', 'enrichment_kit_type', 'manufacturer', 'gene_list', '__str__')
+
+    def get_enrichment_kit_type(self, obj):
+        return obj.get_enrichment_kit_type_display()
+
+    def get___str__(self, obj):
+        return str(obj)
+
+
+class EnrichmentKitSummarySerializer(serializers.ModelSerializer):
+    """ Doesn't return genes (much faster)  """
+    enrichment_kit_type = serializers.SerializerMethodField()
+    manufacturer = serializers.StringRelatedField()
+    __str__ = serializers.SerializerMethodField(read_only=True)
+
+    class Meta:
+        model = EnrichmentKit
+        fields = ('pk', 'name', 'version', 'enrichment_kit_type', 'manufacturer', '__str__')
+
+    def get_enrichment_kit_type(self, obj):
+        return obj.get_enrichment_kit_type_display()
+
+    def get___str__(self, obj):
+        return str(obj)

seqauto/serializers/seqauto_qc_serializers.py

@@ -0,0 +1,69 @@
+from rest_framework import serializers
+
+from genes.serializers import SampleGeneListSerializer, GeneCoverageCollectionSerializer
+from seqauto.models import IlluminaFlowcellQC, QCGeneList, QC, QCGeneCoverage, QCExecSummary, FastQC, SequencingSample
+from seqauto.serializers.sequencing_serializers import SampleSheetLookupSerializer, FastqSerializer, SeqAutoViewMixin, \
+    BamFilePathSerializer, VCFFilePathSerializer, SequencingSampleLookupSerializer
+
+
+class FastQCSerializer(SeqAutoViewMixin, serializers.ModelSerializer):
+    fastq = FastqSerializer()
+
+    class Meta:
+        model = FastQC
+        fields = "__all__"
+
+
+class IlluminaFlowcellQCSerializer(SeqAutoViewMixin, serializers.ModelSerializer):
+    sample_sheet = SampleSheetLookupSerializer()
+
+    class Meta:
+        model = IlluminaFlowcellQC
+        #fields = "__all__"
+        exclude = ("sequencing_run", )  # Already part of sample_sheet
+
+
+class QCSerializer(SeqAutoViewMixin, serializers.ModelSerializer):
+    # Instead of dealing with all the bam/vcf etc - we'll just deal with sequencing_sample and
+    # assume we're using the latest ones associated with that
+    sequencing_sample = serializers.SerializerMethodField(read_only=True)
+    bam_file = BamFilePathSerializer()  # These end up really big, think we just want to pass PKs
+    vcf_file = VCFFilePathSerializer()
+
+    class Meta:
+        model = QC
+        fields = ("sequencing_sample", "bam_file", "vcf_file")
+
+    def get_sequencing_sample(self, instance):
+        try:
+            ss = instance.bam_file.unaligned_reads.sequencing_sample
+            serializer = SequencingSampleLookupSerializer(ss)
+            return serializer.data
+        except SequencingSample.DoesNotExist:
+            return None
+
+
+class QCGeneListSerializer(SeqAutoViewMixin, serializers.ModelSerializer):
+    qc = QCSerializer()
+    sample_gene_list = SampleGeneListSerializer()
+
+    class Meta:
+        model = QCGeneList
+        fields = ("path", "qc", "sample_gene_list")
+
+
+class QCGeneCoverageSerializer(SeqAutoViewMixin, serializers.ModelSerializer):
+    qc = QCSerializer()
+    gene_coverage_collection = GeneCoverageCollectionSerializer()
+
+    class Meta:
+        model = QCGeneCoverage
+        fields = ("path", "qc", "gene_coverage_collection")
+
+
+class QCExecSummarySerializer(SeqAutoViewMixin, serializers.ModelSerializer):
+    qc = QCSerializer()
+
+    class Meta:
+        model = QCExecSummary
+        exclude = ('gene_list', )

seqauto/serializers/seqauto_serializers.py

@@ -0,0 +1,34 @@
+import numpy as np
+from rest_framework import serializers
+
+from genes.serializers import TranscriptSerializer, GeneSymbolSerializer, \
+    TranscriptVersionSerializer
+from seqauto.models import GoldCoverageSummary, GoldReference
+from seqauto.serializers.enrichment_kit_serializers import EnrichmentKitSummarySerializer
+
+
+class GoldReferenceSerializer(serializers.ModelSerializer):
+    enrichment_kit = EnrichmentKitSummarySerializer()
+
+    class Meta:
+        model = GoldReference
+        fields = ('enrichment_kit', 'created')
+
+
+class GoldCoverageSummarySerializer(serializers.ModelSerializer):
+    gold_reference = GoldReferenceSerializer()
+    gene_symbol = GeneSymbolSerializer()
+    transcript = TranscriptSerializer()
+    transcript_version = TranscriptVersionSerializer()
+    standard_error = serializers.SerializerMethodField()
+
+    class Meta:
+        model = GoldCoverageSummary
+        fields = '__all__'
+
+    def get_standard_error(self, obj):
+        """ This can occasionally be NaN which isn't valid JSON """
+        standard_error = obj.standard_error
+        if np.isnan(standard_error):
+            standard_error = -1
+        return standard_error