Merge remote-tracking branch 'origin/main' into simplify-config-files

PMBio · Jun 7, 2024 · 345e4be · 345e4be
2 parents 59de80c + 901ff97
commit 345e4be
Show file tree

Hide file tree

Showing 4 changed files with 6 additions and 7 deletions.
diff --git a/.github/workflows/run-pipeline.yml b/.github/workflows/run-pipeline.yml
@@ -82,7 +82,7 @@ jobs:
             python -m snakemake ${{ (inputs.dry_run && '-n') || '' }} \
             -j 2 --directory ${{inputs.pipeline_directory}} \
             ${{ (endsWith(inputs.pipeline_config, 'ml') && '--configfile')  || '' }}  ${{ inputs.pipeline_config }} \
-            --snakefile ${{inputs.pipeline_file}} --show-failed-logs -F ${{ inputs.pipeline_extra_flags }}
+            --snakefile ${{inputs.pipeline_file}} --show-failed-logs -F -p ${{ inputs.pipeline_extra_flags }}
           shell: micromamba-shell {0}
         - name: Run post pipeline cmd
           if: inputs.postrun_cmd

diff --git a/deeprvat/seed_gene_discovery/config.yaml b/deeprvat/seed_gene_discovery/config.yaml
@@ -60,8 +60,8 @@ variant_file: variants.parquet
 
 data:
     gt_file: genotypes.h5
+    variant_file: variants.parquet
     dataset_config:
-        variant_file: variants.parquet
         phenotype_file: phenotypes.parquet
         standardize_xpheno: False
         y_transformation: quantile_transform

diff --git a/deeprvat/seed_gene_discovery/seed_gene_discovery.py b/deeprvat/seed_gene_discovery/seed_gene_discovery.py
@@ -537,13 +537,12 @@ def make_dataset_(
             dataset = pickle.load(f)
     else:
         logger.info("Instantiating dataset")
-        variant_file = data_config.get(
-            "variant_file", f'{data_config["gt_file"][:-3]}_variants.parquet'
-        )
+
         dataset = DenseGTDataset(
             gt_file=data_config["gt_file"],
             skip_y_na=True,
             skip_x_na=True,
+            variant_file=data_config["variant_file"],
             **data_config["dataset_config"],
         )
         logger.info("Writing pickled data set")

diff --git a/example/config/deeprvat_annotation_config.yaml b/example/config/deeprvat_annotation_config.yaml
@@ -8,12 +8,12 @@ fasta_file_name : hg38.fa
 gtf_file_name : gencode.v44.annotation.gtf.gz
 
 source_variant_file_pattern : test_vcf_data_c{chr}_b{block}
-source_variant_file_type: 'vcf.gz'
+source_variant_file_type: 'bcf'
 
 # comment out / remove to run on all chromosomes
 included_chromosomes : ['21','22']
 
-source_variant_dir : input_dir/vcf
+source_variant_dir : preprocessing_workdir/norm/bcf
 anno_tmp_dir : output_dir/annotations/tmp
 anno_dir : output_dir/annotations