robs analysis updated

scripts/run_robust_analys.sh

@@ -3,11 +3,11 @@
 # RUN_ID="run_$(date +%Y-%m-%d_%H-%M-%S)"
 
 RUN_ID="robust_analy"
-# resources_dir="resources"
-# publish_dir="output/${RUN_ID}"
+resources_dir="resources"
+publish_dir="output/${RUN_ID}"
 
-resources_dir="s3://openproblems-data/resources/grn"
-publish_dir="s3://openproblems-data/resources/grn/results/${RUN_ID}"
+# resources_dir="s3://openproblems-data/resources/grn"
+# publish_dir="s3://openproblems-data/resources/grn/results/${RUN_ID}"
 
 grn_models_folder="${resources_dir}/grn_models"
 

src/api/comp_metric.yaml

@@ -15,7 +15,7 @@ functionality:
       default: resources/grn-benchmark/perturbation_data.h5ad
     - name: --prediction
       __merge__: file_prediction.yaml
-      required: false
+      required: true
       direction: input
       default: resources/grn_models/collectri.csv
     - name: --score

src/methods/multi_omics/scenicplus/config.vsh.yaml

@@ -11,10 +11,9 @@ functionality:
       GRN inference using scenicplus
     documentation_url: https://scenicplus.readthedocs.io/en/latest/human_cerebellum.html
   arguments:
-    - name: --cistopic_out
+    - name: --cistopic_object
       type: file
-      required: false
-      direction: input
+      direction: output
   resources:
     - type: python_script
       path: script.py

src/methods/multi_omics/scenicplus/script.py

@@ -21,13 +21,14 @@
 par = {
   'multiomics_rna': 'resources/grn-benchmark/multiomics_rna.h5ad',
   'multiomics_atac': 'resources/grn-benchmark/multiomics_atac.h5ad',
-  'cistopic_out': 'output/pycistopic',
   'temp_dir': 'output/scenicplus',
   'prediction': 'output/prediction.csv',
 }
 ## VIASH END
 
 work_dir = par['temp_dir']
+par['cistopic_out'] = f'{work_dir}/cistopic_out'
+par['cistopic_object'] = os.path.join(par['cistopic_out'], f'cistopic_object_with_model.pkl')
 os.makedirs(os.path.join(work_dir, 'scRNA'), exist_ok=True)
 
 # Download databases
@@ -96,6 +97,7 @@ def download_checksum(url: str, filepath: str) -> str:
 with open(os.path.join(par['cistopic_out'], f'candidate_enhancers/markers_dict.pkl'), 'rb') as f:
     markers_dict = pickle.load(f)
 
+
 # Convert to dictionary of pyrange objects
 region_sets = {}
 region_sets['topics_otsu'] = {}
@@ -126,7 +128,7 @@ def download_checksum(url: str, filepath: str) -> str:
     settings = yaml.safe_load(f)
 
 # Update settings: indicate locations of input files
-settings['input_data']['cisTopic_obj_fname'] = os.path.join(par['cistopic_out'], f'cistopic_object_with_model.pkl')
+settings['input_data']['cisTopic_obj_fname'] = par['cistopic_object']
 settings['input_data']['GEX_anndata_fname'] = os.path.join(work_dir, 'rna.h5ad')
 settings['input_data']['region_set_folder'] = os.path.join(par['cistopic_out'], 'region_sets')
 settings['input_data']['ctx_db_fname'] = rankings_db
@@ -152,3 +154,10 @@ def download_checksum(url: str, filepath: str) -> str:
 # Make sure the file is properly formatted, and re-format it if needed
 filepath = os.path.join(work_dir, 'tf_to_gene_adj.tsv')
 shutil.copyfile(filepath, par['prediction'])
+
+
+# cistopic_obj = pickle.load(os.path.join(par['cistopic_out'], f'cistopic_object_with_model.pkl'))
+# # get cell topic association 
+# cell_topic = cistopic_obj.selected_model.cell_topic.T
+# cell_names = cistopic_obj.cell_data.obs_id.values
+# cell_topic.index = cell_names

src/methods/multi_omics/scenicplus_ns/config.vsh.yaml

@@ -7,6 +7,10 @@ functionality:
   info:
     label: grn_inference_scenicplus
     summary: "Infers GRNs from multiomics data using scenicplus."
+  arguments:
+    - name: --cistopic_object
+      type: file
+      direction: output
 
   resources:
     - type: nextflow_script

src/methods/multi_omics/scenicplus_ns/main.nf

@@ -11,10 +11,10 @@ workflow run_wf {
               temp_dir: "temp_dir",
               num_workers: "num_workers"
               ],
-      toState: [prediction:"prediction"]
+      toState: [prediction:"prediction", cistopic_object:"cistopic_object"]
     )
 
-    | setState(["prediction"])
+    | setState(["prediction", "cistopic_object"])
 
   emit:
   output_ch

src/methods/multi_omics/scenicplus_ns/run.sh

@@ -31,14 +31,14 @@ HERE
 #   -params-file params/${RUN_ID}.yaml
 
 
-# ./tw-windows-x86_64.exe launch `
-#       https://github.com/openproblems-bio/task_grn_benchmark.git `
-#       --revision build/main `
-#       --pull-latest `
-#       --main-script target/nextflow/workflows/run_grn_inference/main.nf `
-#       --workspace 53907369739130 `
-#       --compute-env 6TeIFgV5OY4pJCk8I0bfOh `
-#       --params-file ./params/celloracle_test.yaml `
-#       --config src/common/nextflow_helpers/labels_tw.config
+./tw-windows-x86_64.exe launch `
+      https://github.com/openproblems-bio/task_grn_benchmark.git `
+      --revision build/main `
+      --pull-latest `
+      --main-script target/nextflow/workflows/grn_inference_scenicplus/main.nf `
+      --workspace 53907369739130 `
+      --compute-env 6TeIFgV5OY4pJCk8I0bfOh `
+      --params-file ./params/scenicplus.yaml `
+      --config src/common/nextflow_helpers/labels_tw.config
 
 

src/workflows/run_robustness_analysis/main.nf

@@ -43,7 +43,7 @@ workflow run_wf {
       // use 'fromState' to fetch the arguments the component requires from the overall state
       fromState: [
         perturbation_data: "perturbation_data",
-        prediction_n: "prediction",
+        prediction: "prediction_n",
         layer: "layer", 
         subsample: "subsample",
         reg_type: "reg_type",