only use project assembly as target if there's exactly one

eva_submission/eload_ingestion.py

@@ -423,8 +423,9 @@ def _get_target_assembly(self):
         4. If there is exactly one alternate taxonomy, check Ensembl (main then rapid release) for its supported
            assembly. If present use that, and insert into supported assembly table.
         5. If we still do not have a target assembly (i.e. project taxonomy is not supported by Ensembl AND alternate
-           taxonomy cannot be determined or is not supported by Ensembl), then choose one of the project's
-           assemblies and insert into the supported assembly table.
+           taxonomy cannot be determined or is not supported by Ensembl), and the project has exactly one assembly,
+           then choose that one and insert into the supported assembly table.
+        6. Otherwise return None (which will skip clustering)
         """
         if self.taxonomy == 9606:
             self.info('No remapping or clustering for human studies')
@@ -435,8 +436,10 @@ def _get_target_assembly(self):
             target_assembly = self._get_supported_assembly_from_evapro(alt_tax_id) or \
                               self._insert_new_supported_asm_from_ensembl(alt_tax_id)
         if target_assembly is None:
-            # TODO any particular one?
-            target_assembly = list(self.assembly_accessions)[0]
+            if len(self.assembly_accessions) == 1:
+                target_assembly = list(self.assembly_accessions)[0]
+                add_to_supported_assemblies(self.metadata_connection_handle, source_of_assembly='EVA',
+                                            target_assembly=target_assembly, taxonomy_id=self.taxonomy)
         return target_assembly
 
     def _get_alt_tax_id(self):

requirements.txt

@@ -1,6 +1,6 @@
 cached-property
 cerberus
-ebi-eva-common-pyutils>=0.5.6
+ebi-eva-common-pyutils==0.5.8.dev0
 eva-vcf-merge>=0.0.6
 humanize
 lxml

tests/test_eload_ingestion.py

@@ -15,10 +15,13 @@ def default_db_results_for_metadata_load():
         [(391,)]  # Check the assembly_set_id in update_assembly_set_in_analysis
     ]
 
+
 def default_db_results_for_target_assembly():
     return [
         [('GCA_999')]
     ]
+
+
 def default_db_results_for_accession():
     browsable_files = [(1, 'ERA', 'filename_1', 'PRJ', 123), (2, 'ERA', 'filename_1', 'PRJ', 123)]
     return [
@@ -44,7 +47,6 @@ def default_db_results_for_clustering():
 def default_db_results_for_ingestion():
     return (
             default_db_results_for_metadata_load()
-            + default_db_results_for_target_assembly()
             + default_db_results_for_accession()
             + default_db_results_for_clustering()
             + default_db_results_for_variant_load()
@@ -207,7 +209,7 @@ def test_ingest_accession(self):
             m_get_vep_versions.return_value = (100, 100)
             m_get_species.return_value = 'homo_sapiens'
             m_post.return_value.text = self.get_mock_result_for_ena_date()
-            m_get_results.side_effect = default_db_results_for_target_assembly() + default_db_results_for_accession()
+            m_get_results.side_effect = default_db_results_for_accession()
             self.eload.ingest(
                 instance_id=1,
                 tasks=['accession']
@@ -469,7 +471,6 @@ def test_resume_when_step_fails(self):
             m_get_species.return_value = 'homo_sapiens'
             m_post.return_value.text = self.get_mock_result_for_ena_date()
             m_get_results.side_effect = default_db_results_for_metadata_load() \
-                                        + default_db_results_for_target_assembly()\
                                         + default_db_results_for_ingestion()
 
             m_run_command.side_effect = [
@@ -535,9 +536,10 @@ def test_resume_with_tasks(self):
             m_get_vep_versions.return_value = (100, 100)
             m_get_species.return_value = 'homo_sapiens'
             m_post.return_value.text = self.get_mock_result_for_ena_date()
-            m_get_results.side_effect = (default_db_results_for_target_assembly()
-                    + default_db_results_for_variant_load() + default_db_results_for_target_assembly()
-                    + default_db_results_for_accession() + default_db_results_for_variant_load()
+            m_get_results.side_effect = (
+                    default_db_results_for_variant_load()
+                    + default_db_results_for_accession()
+                    + default_db_results_for_variant_load()
             )
 
             m_run_command.side_effect = [
@@ -568,3 +570,7 @@ def test_resume_with_tasks(self):
                                                                     'nextflow_dir')
             assert new_accession_nextflow_dir == self.eload.nextflow_complete_value
             assert not os.path.exists(accession_nextflow_dir)
+
+    def test_get_target_assembly(self):
+        # TODO mock only the db, run everything else
+        ...