refactor: delete unnecessary config files (#647)

* refactor: remove ot_pics * refactor: gwas_catalog_sumstat_preprocess config removed * refactor: ot_finngen_studies removed * refactor: ot_finngen_studies removed * refactor: window_based_clumping cleanup
opentargets · Jun 18, 2024 · dc70bd8 · dc70bd8
1 parent d796b68
commit dc70bd8
Show file tree

Hide file tree

Showing 10 changed files with 10 additions and 33 deletions.
diff --git a/config/step/ot_finngen_studies.yaml b/config/step/ot_finngen_studies.yaml
diff --git a/config/step/ot_finngen_sumstat_preprocess.yaml b/config/step/ot_finngen_sumstat_preprocess.yaml
diff --git a/config/step/ot_gwas_catalog_sumstat_preprocess.yaml b/config/step/ot_gwas_catalog_sumstat_preprocess.yaml
diff --git a/config/step/ot_pics.yaml b/config/step/ot_pics.yaml
diff --git a/config/step/ot_window_based_clumping.yaml b/config/step/ot_window_based_clumping.yaml
diff --git a/src/airflow/dags/finngen_harmonisation.py b/src/airflow/dags/finngen_harmonisation.py
@@ -1,4 +1,5 @@
 """Airflow DAG for the harmonisation part of the pipeline."""
+
 from __future__ import annotations
 
 import re
@@ -53,7 +54,7 @@ def submit_jobs(**kwargs: Any) -> None:
             print("Submitting job for study: ", study_id)  # noqa: T201
             common.submit_pyspark_job_no_operator(
                 cluster_name=CLUSTER_NAME,
-                step_id="ot_finngen_sumstat_preprocess",
+                step_id="finngen_sumstat_preprocess",
                 other_args=[
                     f"step.raw_sumstats_path=gs://{SUMMARY_STATS_BUCKET_NAME}/{input_path}",
                     f"step.out_sumstats_path={SUMSTATS_PARQUET}/{study_id}.parquet",

diff --git a/src/airflow/dags/finngen_preprocess.py b/src/airflow/dags/finngen_preprocess.py
@@ -53,7 +53,7 @@
     ) as finngen_summary_stats_preprocess:
         study_index = common.submit_step(
             cluster_name=CLUSTER_NAME,
-            step_id="ot_finngen_studies",
+            step_id="finngen_studies",
             task_id="finngen_studies",
             other_args=[
                 f"step.finngen_study_index_out={STUDY_INDEX}",

diff --git a/src/airflow/dags/gwas_catalog_harmonisation.py b/src/airflow/dags/gwas_catalog_harmonisation.py
@@ -1,4 +1,5 @@
 """Airflow DAG for the harmonisation part of the pipeline."""
+
 from __future__ import annotations
 
 import re
@@ -99,7 +100,7 @@ def submit_jobs(**kwargs: Any) -> None:
             print("Submitting job for study: ", study_id)  # noqa: T201
             common.submit_pyspark_job_no_operator(
                 cluster_name=CLUSTER_NAME,
-                step_id="ot_gwas_catalog_sumstat_preprocess",
+                step_id="gwas_catalog_sumstat_preprocess",
                 other_args=[
                     f"step.raw_sumstats_path=gs://{SUMMARY_STATS_BUCKET_NAME}/{input_path}",
                     f"step.out_sumstats_path=gs://{SUMMARY_STATS_BUCKET_NAME}/{HARMONISED_SUMMARY_STATISTICS_PREFIX}/{study_id}.parquet",

diff --git a/src/airflow/dags/gwas_catalog_preprocess.py b/src/airflow/dags/gwas_catalog_preprocess.py
@@ -1,4 +1,5 @@
 """Airflow DAG for the preprocessing of GWAS Catalog's harmonised summary statistics and curated associations."""
+
 from __future__ import annotations
 
 from pathlib import Path
@@ -131,7 +132,7 @@ def upload_harmonized_study_list(
         # Do PICS based finemapping:
         curation_pics = common.submit_step(
             cluster_name=CLUSTER_NAME,
-            step_id="ot_pics",
+            step_id="pics",
             task_id="catalog_curation_pics",
             other_args=[
                 f"step.study_locus_ld_annotated_in={CURATED_LD_CLUMPED}",
@@ -167,7 +168,7 @@ def upload_harmonized_study_list(
         # Run window-based clumping:
         summary_stats_window_based_clumping = common.submit_step(
             cluster_name=CLUSTER_NAME,
-            step_id="ot_window_based_clumping",
+            step_id="window_based_clumping",
             task_id="catalog_sumstats_window_clumping",
             other_args=[
                 f"step.summary_statistics_input_path=gs://{GWAS_CATALOG_BUCKET_NAME}/{HARMONISED_SUMSTATS_PREFIX}",
@@ -191,7 +192,7 @@ def upload_harmonized_study_list(
         # Run PICS finemapping:
         summary_stats_pics = common.submit_step(
             cluster_name=CLUSTER_NAME,
-            step_id="ot_pics",
+            step_id="pics",
             task_id="catalog_sumstats_pics",
             other_args=[
                 f"step.study_locus_ld_annotated_in={LD_BASED_CLUMPED}",

diff --git a/src/gentropy/config.py b/src/gentropy/config.py
@@ -380,7 +380,7 @@ class WindowBasedClumpingStepConfig(StepConfig):
     )
     summary_statistics_input_path: str = MISSING
     study_locus_output_path: str = MISSING
-    gwas_significance: float = 5e-8
+    gwas_significance: float = 1e-8
     distance: int = 500_000
     collect_locus: bool = False
     collect_locus_distance: int = 500_000