[long-reads]: add customised input option

Author: msarkis-icr

custom_files/pipeline_project_map.json

@@ -0,0 +1,9 @@
+epi2me-human-variation:
+  nf-long-reads:
+    is_inputType_path: False
+epi2me-somatic-variation:
+  mopopgen-support:
+    is_inputType_path: True
+nfcore-rnaseq:
+  nf-tp53:
+    is_inputType_path: True

custom_files/pipeline_project_map_specifications.yaml

@@ -1,4 +1,5 @@
 import streamlit as st
+import yaml
 
 import tabs.tab_command as tt
 from pipeline_project_map import map_pipeline_project
@@ -33,29 +34,48 @@ def reset_button_state():
 OUTPUT_DIR = ss_values["OUTPUT_DIR"]
 run_pipeline_clicked = ss_values["run_pipeline_clicked"]
 button_clicked = ss_values["button_clicked"]
+custom_sample_list = ss_values["custom_sample_list"]  # only availanle if custom sample is selected
 
-samples = ["all", "demo"]  # , "customised"]
+samples = ["test", "all", "demo", "customised"]
 
 # Create the selectbox and update session state
-options = ["select"] + list(map_pipeline_project.keys())
-index = options.index(PIPELINE)
-PIPELINE = st.selectbox("Select a pipeline", options=options, index=index)  # , key="PIPELINE")
+pipeline_options = ["select"] + list(map_pipeline_project.keys())
+index = pipeline_options.index(PIPELINE)
+PIPELINE = st.selectbox("Select a pipeline", options=pipeline_options, index=index)  # , key="PIPELINE")
 # adding "select" as the first and default choice
 if PIPELINE != "select":
+    project_options = list(map_pipeline_project[PIPELINE].keys())
+
     PROJECT = st.selectbox(
         "Select your project",
-        options=map_pipeline_project[PIPELINE],
-        index=map_pipeline_project[PIPELINE].index(PIPELINE) if PIPELINE in map_pipeline_project[PIPELINE] else 0,
+        options=project_options,
+        index=project_options.index(PIPELINE) if PIPELINE in project_options else 0,
         on_change=reset_button_state,
     )
-
+    # samples here depend on the pipeline: human variation requires bam files, rnaseq requires a samplesheet
     SAMPLE = st.selectbox(
         "Select your samples",
         options=samples,
         index=samples.index(SAMPLE) if SAMPLE in samples else 0,
         on_change=reset_button_state,
     )
 
+    # If "customised" is selected, show additional input
+    if SAMPLE == "customised":
+        is_input_type_path = map_pipeline_project[PIPELINE][PROJECT]["is_inputType_path"]
+        msg = "Enter your sample names (comma-separated)"
+        if is_input_type_path:
+            msg = "Enter path to samplesheet"
+        custom_samples = st.text_input(msg, key="custom_samples", value=",".join(custom_sample_list))
+
+        # Optionally process the input into a list
+        if custom_samples:
+            custom_sample_list = [s.strip() for s in custom_samples.split(",") if s.strip()]
+
+            st.write("Your custom samples:", custom_sample_list)
+            ss_set("custom_sample_list", custom_sample_list)
+
+    # for tp53, we need to provide a sample sheet while for human-variation a simple sample list if enough
 WORK_DIR = st.text_input("Working directory", value=SCRATCH)
 OUTPUT_DIR = st.text_input("Output directory", value=SCRATCH)
 
@@ -69,6 +89,7 @@ def reset_button_state():
         selected_samples=SAMPLE,
         work_dir=WORK_DIR,
         output_dir=OUTPUT_DIR,
+        custom_sample_list=custom_sample_list,
     )
     save_in_ss(
         {
@@ -89,6 +110,7 @@ def reset_button_state():
             "OUTPUT_DIR": OUTPUT_DIR,
             "run_pipeline_clicked": run_pipeline_clicked,
             "button_clicked": button_clicked,
+            "custom_sample_list": custom_sample_list,
         }
     )
 else:

pages/run_pipeline.py

@@ -1,11 +1,6 @@
-import json
+import yaml
 
+file_path = "custom_files/pipeline_project_map_specifications.yaml"
 
-def load_json_to_dict(file_path):
-    with open(file_path, "r") as file:
-        data = json.load(file)
-    return data
-
-
-file_path = "custom_files/pipeline_project_map.json"
-map_pipeline_project = load_json_to_dict(file_path)
+with open(file_path, "r") as f:
+    map_pipeline_project = yaml.safe_load(f)

pipeline_project_map.py

@@ -9,13 +9,13 @@ def get_path_to_script(selected_pipeline, selected_project, selected="all"):
     script_mapping = {
         "all": "launch_samples.sh",
         "demo": "launch_demo.sh",
-        "single": "launch_sample_analysis.sh",  # not yet supported
+        "customised": "launch_samples.sh",
     }
 
     if selected in script_mapping:
         return os.path.join(base_path, script_mapping[selected])
 
-    raise ValueError(f"Invalid selection '{selected}'. Only 'all' and 'demo' are supported.")
+    raise ValueError(f"Invalid selection '{selected}'. Only 'customised' and 'demo' are supported.")
 
 
 # launch command based on the project
@@ -27,6 +27,7 @@ def pipe_cmd(
     selected_samples="all",
     work_dir="work",
     output_dir="output",
+    custom_sample_list=[],
 ):
     def get_pipeline_command():
         """Generate the pipeline execution command based on the sample selection."""
@@ -42,7 +43,7 @@ def get_pipeline_command():
 
         if selected_samples == "demo":
             cmd_pipeline += f"sbatch  -o {log_out} -e {log_err} {path_to_script} {work_dir} {output_dir}"
-        elif selected_samples == "all":
+        elif selected_samples == "all":  # this has no more sense since we have to specify the sample name index
             cmd_pipeline += f"sbatch  -o {log_out} -e {log_err} {path_to_script} --work-dir {work_dir} --outdir {output_dir}"
             ##./your_script.sh --env "/my/custom/env" --work-dir "my_work" --outdir "my_output" --config "my_config" --params "parans.json"
             # Usage:
@@ -52,7 +53,14 @@ def get_pipeline_command():
             #     --env "/data/rds/DIT/SCICOM/SCRSE/shared/conda/nextflow_env" \
             #     --params "/data/params/parameters.json" \
             #     --config "custom_config.config"
+        elif selected_samples == "customised":
+            if not len(custom_sample_list):
+                print("custom_sample_list cannot be empty")
 
+            tab_separated_string = "\t".join(custom_sample_list)
+            cmd_pipeline += f"sbatch  -o {log_out} -e {log_err} {path_to_script} --work-dir {work_dir} --outdir {output_dir} --samples {tab_separated_string}"
+        elif selected_samples == "test":
+            cmd_pipeline += f"sbatch  -o {log_out} -e {log_err} /data/scratch/DCO/DIGOPS/SCIENCOM/msarkis/NF-project-configurations/test.sh --work-dir {work_dir} --outdir {output_dir}"
         return cmd_pipeline.strip()
 
     # Command mappings

shared/command_helper.py

@@ -17,3 +17,4 @@ keys_defaults:
   group_selection: "Select an option"
   WORK_DIR: None
   OUTPUT_DIR: None
+  custom_sample_list: []

shared/ss_defaults.yaml

@@ -24,6 +24,7 @@
 WORKDIR = ss_values["WORK_DIR"]
 OUTPUT_DIR = ss_values["OUTPUT_DIR"]
 run_pipeline_clicked = ss_values["run_pipeline_clicked"]
+custom_sample_list = ss_values["custom_sample_list"]  # only availanle if custom sample is selected
 
 
 def display_log(title, log_path, output_container):
@@ -37,7 +38,16 @@ def display_log(title, log_path, output_container):
         st.error(f"Failed to read {title.lower()} log: {e}")
 
 
-def tab(username, MY_SSH, selected_pipeline, selected_project, selected_samples="all", work_dir="work", output_dir="output"):
+def tab(
+    username,
+    MY_SSH,
+    selected_pipeline,
+    selected_project,
+    selected_samples="all",
+    work_dir="work",
+    output_dir="output",
+    custom_sample_list=[],
+):
     JOB_ID = ss_values["JOB_ID"]
 
     cols = st.columns([1, 1, 1])
@@ -58,6 +68,7 @@ def run_nextflow():
             selected_samples=selected_samples,
             output_dir=output_dir,
             work_dir=work_dir,
+            custom_sample_list=custom_sample_list,
         )
         st.code(cmd_pipeline)
         _dict = MY_SSH.run_cmd(cmd_pipeline)

tabs/tab_command.py

@@ -80,7 +80,6 @@ def create_radio(GROUPS):
         OK, MY_SSH, msg, GROUPS = handle_login(server, sftp_server, username, password)
         if OK:
             st.success(msg)
-            create_radio(GROUPS)
         else:
             st.error(msg)