Share a single GemmaApi instance

Author: arteymix

rnaseq_pipeline/gemma.py

@@ -71,17 +71,15 @@ def platforms(self, experiment_id):
     def quantitation_types(self, experiment_id):
         return self._query_api(join('datasets', experiment_id, 'quantitationTypes'))
 
+gemma_api = GemmaApi()
+
 class GemmaTaskMixin(luigi.Task):
     experiment_id = luigi.Parameter()
 
-    def __init__(self, *kwargs, **kwds):
-        super().__init__(*kwargs, **kwds)
-        self._gemma_api = GemmaApi()
-
     @property
     def dataset_info(self):
         if not hasattr(self, '_dataset_info'):
-            data = self._gemma_api.datasets(self.experiment_id)
+            data = gemma_api.datasets(self.experiment_id)
             if not data:
                 raise RuntimeError('Could not retrieve Gemma dataset with short name {}.'.format(self.experiment_id))
             self._dataset_info = data[0]
@@ -112,7 +110,7 @@ def reference_id(self):
         try:
             if self.assay_type == GemmaAssayType.BULK_RNA_SEQ:
                 return {'human': cfg.human_reference_id, 'mouse': cfg.mouse_reference_id, 'rat': cfg.rat_reference_id}[
-                self.taxon]
+                    self.taxon]
             elif self.assay_type == GemmaAssayType.SINGLE_CELL_RNA_SEQ:
                 return {'human': cfg.human_single_cell_reference_id, 'mouse': cfg.mouse_single_cell_reference_id,
                         'rat': cfg.rat_single_cell_reference_id}[
@@ -150,7 +148,7 @@ def assay_type(self):
                           'http://www.ebi.ac.uk/efo/EFO_0005684']
         fac_sorted_uri = 'http://www.ebi.ac.uk/efo/EFO_0009108'
 
-        annotations = self._gemma_api.dataset_annotations(self.experiment_id)
+        annotations = gemma_api.dataset_annotations(self.experiment_id)
         fac_sorted = any(annotation['classUri'] == assay_type_class_uri and annotation['termUri'] == fac_sorted_uri
                          for annotation in annotations)
         for annotation in annotations:

rnaseq_pipeline/sources/gemma.py

@@ -1,12 +1,12 @@
 import logging
 
 import luigi
-
 from bioluigi.tasks.utils import DynamicTaskWithOutputMixin, DynamicWrapperTask
+
 from .geo import DownloadGeoSample
 from .sra import DownloadSraExperiment
 from ..config import Config
-from ..gemma import GemmaApi
+from ..gemma import gemma_api
 
 logger = logging.getLogger(__name__)
 
@@ -21,12 +21,8 @@ class DownloadGemmaExperiment(DynamicTaskWithOutputMixin, DynamicWrapperTask):
     """
     experiment_id: str = luigi.Parameter()
 
-    def __init__(self, *kwargs, **kwds):
-        super().__init__(*kwargs, **kwds)
-        self._gemma_api = GemmaApi()
-
     def run(self):
-        data = self._gemma_api.samples(self.experiment_id)
+        data = gemma_api.samples(self.experiment_id)
         download_sample_tasks = []
         for sample in data:
             accession = sample['accession']['accession']
@@ -41,4 +37,4 @@ def run(self):
             else:
                 logger.warning('Downloading %s from %s is not supported.', accession, external_database)
                 continue
-        yield download_sample_tasks
+        yield download_sample_tasks

rnaseq_pipeline/targets.py

@@ -8,7 +8,7 @@
 
 import luigi
 
-from .gemma import GemmaApi
+from .gemma import gemma_api
 
 logger = logging.getLogger(__name__)
 
@@ -52,14 +52,13 @@ def __init__(self,
         self.dataset = dataset
         self.quantitation_type = quantitation_type
         self.vector_type = vector_type
-        self._gemma_api = GemmaApi()
 
     def exists(self):
         return any(
             (quantitation_type['id'] == self.quantitation_type or quantitation_type['name'] == self.quantitation_type
              if self.quantitation_type else quantitation_type['isPreferred'])
             and (quantitation_type['vectorType'] == self.vector_type.value if self.vector_type else True)
-            for quantitation_type in self._gemma_api.quantitation_types(self.dataset))
+            for quantitation_type in gemma_api.quantitation_types(self.dataset))
 
     def __repr__(self):
         return f'GemmaDatasetQuantitationType(dataset={self.dataset}, quantitation_type={self.quantitation_type}, vector_type={self.vector_type})'
@@ -71,10 +70,9 @@ class GemmaDatasetHasBatch(luigi.Target):
 
     def __init__(self, dataset_short_name):
         self.dataset_short_name = dataset_short_name
-        self._gemma_api = GemmaApi()
 
     def exists(self):
-        return self._gemma_api.dataset_has_batch(self.dataset_short_name)
+        return gemma_api.dataset_has_batch(self.dataset_short_name)
 
 class ExpirableLocalTarget(luigi.LocalTarget):
     """

rnaseq_pipeline/tasks.py

@@ -23,7 +23,7 @@
 from luigi.util import requires
 
 from rnaseq_pipeline.config import Config
-from .gemma import GemmaAssayType, GemmaTaskMixin, GemmaConfig
+from .gemma import GemmaAssayType, GemmaTaskMixin, GemmaConfig, gemma_api
 from .gemma import GemmaCliTask
 from .sources.arrayexpress import DownloadArrayExpressSample, DownloadArrayExpressExperiment
 from .sources.gemma import DownloadGemmaExperiment
@@ -819,7 +819,7 @@ def run(self):
         for split_id in range(self.num_splits):
             split_id = self.experiment_id + '.' + str(split_id + 1)
             logger.info('Reorganizing data for %s.', split_id)
-            for sample in self._gemma_api.samples(split_id):
+            for sample in gemma_api.samples(split_id):
                 sample_id = sample['accession']['accession']
                 for d in dirs_to_relocate:
                     for sample_file in iglob(join(d, self.experiment_id, sample_id)):