scglue ns added

scripts/run_grn_evaluation_tw.sh

@@ -51,7 +51,7 @@ done
 grn_name="negative_control"
 cat >> $param_file << HERE
   - id: ${layer}_${grn_name}
-    perturbation_data: ${perturbation_data}
+    perturbation_data: ${resources_dir}/grn-benchmark/perturbation_data.h5ad
     layer: ${layer}
     reg_type: $reg_type
     method_id: $grn_name

src/methods/multi_omics/scglue/main.py

@@ -245,8 +245,6 @@ def cis_inference(par):
         print("pyscenic ctx failed with return code", result.returncode)
 
 def main(par):
-
-
     os.makedirs(par['temp_dir'], exist_ok=True)
     print('Reading input files', flush=True)
     rna = ad.read_h5ad(par['multiomics_rna'])
@@ -258,15 +256,15 @@ def main(par):
     training(par)
     cis_inference(par)
     print('Curate predictions', flush=True)
-    df = pd.read_csv(
+    pruned_grn = pd.read_csv(
         f"{par['temp_dir']}/pruned_grn.csv", header=None, skiprows=3,
         usecols=[0, 8], names=["TF", "targets"]
     )
 
     tfs_list = []
     target_list = []
     weight_list = []
-    for i, (tf, targets) in df.iterrows():
+    for i, (tf, targets) in pruned_grn.iterrows():
         for target, weight in literal_eval(targets):
             tfs_list.append(tf)
             target_list.append(target)
@@ -275,10 +273,5 @@ def main(par):
     scglue_grn.weight = scglue_grn.weight.astype(float)
     scglue_grn = scglue_grn.drop_duplicates().reset_index(drop=True)
 
-
-    # scglue_grn = pd.DataFrame(
-    # data = {'source':['tf1'], 'target':['g1'], 'weight':[1]}
-    # )
-
     return scglue_grn
 

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

@@ -0,0 +1,29 @@
+
+__merge__: ../../../api/comp_method.yaml
+
+functionality:
+  name: grn_inference_scglue
+  namespace: "workflows"
+  info:
+    label: grn_inference_scglue
+    summary: "Infers GRNs from multiomics data using scglue."
+  arguments:
+    - name: --annotation_file
+      type: file
+      required: true
+      direction: input
+    - name: --motif_file
+      type: file
+      required: true
+      direction: input
+  resources:
+    - type: nextflow_script
+      path: main.nf
+      entrypoint: run_wf
+  dependencies:
+    - name: grn_methods/scglue 
+
+platforms:
+  - type: nextflow
+    directives:
+      label: [ midtime, midmem, lowcpu ]

src/methods/multi_omics/scglue_ns/main.nf

@@ -0,0 +1,23 @@
+workflow run_wf {
+  take:
+  input_ch
+
+  main:
+  output_ch = input_ch
+
+    | scglue.run(
+      fromState: [multiomics_rna: "multiomics_rna",
+              multiomics_atac: "multiomics_atac",
+              temp_dir: "temp_dir",
+              num_workers: "num_workers",
+              annotation_file: "annotation_file",
+              motif_file: "motif_file"
+              ],
+      toState: [prediction:"prediction"]
+    )
+
+    | setState(["prediction"])
+
+  emit:
+  output_ch
+}

src/methods/multi_omics/scglue_ns/run.sh

@@ -0,0 +1,26 @@
+#!/bin/bash
+
+# RUN_ID="run_$(date +%Y-%m-%d_%H-%M-%S)"
+submit=false
+read_results=true
+
+RUN_ID="scglue"
+resources_dir="s3://openproblems-data/resources/grn/"
+publish_dir="s3://openproblems-data/resources/grn/results/${RUN_ID}"
+
+num_workers=20
+
+param_file="./params/${RUN_ID}.yaml"
+
+cat > $param_file << HERE
+param_list:
+  - id: ${RUN_ID}
+    multiomics_rna: ${resources_dir}/grn-benchmark/multiomics_rna.h5ad
+    multiomics_atac: ${resources_dir}/grn-benchmark/multiomics_atac.h5ad
+    annotation_file: ${resources_dir}/supplements/gencode.v45.annotation.gtf.gz
+    motif_file: ${resources_dir}/supplements/JASPAR2022-hg38.bed.gz
+    num_workers: $num_workers
+    temp_dir: ./tmp/grn
+output_state: "state.yaml"
+publish_dir: "$publish_dir"
+HERE

src/metrics/regression_2/config.vsh.yaml

@@ -2,7 +2,7 @@ __merge__: ../../api/comp_metric.yaml
 functionality:
   name: regression_2
   info:
-    label: Regression 2
+    label: regression_2
     summary: Calculates R2 score for regression 2
     description: |
       Calculates R2 score using regression approach 2.