Change config handling

.github/workflows/_run_tests.yml

@@ -34,6 +34,11 @@ jobs:
     - name: Conda info
       shell: bash -l {0}
       run: conda info
+    - name: Install mono
+      shell: bash -l {0}
+      run: |
+        conda install mono
+
     - name: Perform pip installation with all stable dependencies
       shell: bash -l {0}
       run: |

alphadia/cli.py

@@ -16,14 +16,18 @@
 
 import alphadia
 from alphadia import utils
+from alphadia.constants.keys import ConfigKeys
 from alphadia.exceptions import CustomError
 from alphadia.search_plan import SearchPlan
 from alphadia.workflow import reporting
 
 logger = logging.getLogger()
 
+epilog = "Parameters passed via CLI will overwrite parameters from config file (except for  '--file': will be merged)."
 
-parser = argparse.ArgumentParser(description="Search DIA experiments with alphaDIA")
+parser = argparse.ArgumentParser(
+    description="Search DIA experiments with alphaDIA", epilog=epilog
+)
 parser.add_argument(
     "--version",
     "-v",
@@ -32,17 +36,19 @@
 )
 parser.add_argument(
     "--output",
+    "--output-directory",
     "-o",
     type=str,
-    help="Output directory",
+    help="Output directory.",
     nargs="?",
     default=None,
 )
 parser.add_argument(
     "--file",
+    "--raw-path",
     "-f",
     type=str,
-    help="Raw data input files.",
+    help="Path to raw data input file. Can be passed multiple times.",
     action="append",
     default=[],
 )
@@ -58,29 +64,31 @@
     "--regex",
     "-r",
     type=str,
-    help="Regex to match raw files in directory.",
+    help="Regex to match raw files in 'directory'.",
     nargs="?",
     default=".*",
 )
 parser.add_argument(
     "--library",
+    "--library-path",
     "-l",
     type=str,
-    help="Spectral library.",
+    help="Path to spectral library file.",
     nargs="?",
     default=None,
 )
 parser.add_argument(
     "--fasta",
-    help="Fasta file(s) used to generate or annotate the spectral library.",
+    "--fasta-path",
+    help="Path to fasta file used to generate or annotate the spectral library. Can be passed multiple times.",
     action="append",
     default=[],
 )
 parser.add_argument(
     "--config",
     "-c",
     type=str,
-    help="Config yaml which will be used to update the default config.",
+    help="Path to config yaml file which will be used to update the default config.",
     nargs="?",
     default=None,
 )
@@ -93,14 +101,15 @@
 parser.add_argument(
     "--config-dict",
     type=str,
-    help="Python Dict which will be used to update the default config.",
+    help="Python dictionary which will be used to update the default config.",
     nargs="?",
     default="{}",
 )
 parser.add_argument(
-    "--quant-dir",
+    "--quant-dir",  # TODO deprecate
+    "--quant-directory",
     type=str,
-    help="Directory to save the quantification results (psm & frag parquet files) to be reused in a distributed search",
+    help="Directory to save the quantification results (psm & frag parquet files) to be reused in a distributed search.",
     nargs="?",
     default=None,
 )
@@ -159,11 +168,11 @@ def parse_output_directory(args: argparse.Namespace, config: dict) -> str:
     """
 
     output_directory = None
-    if "output_directory" in config:
+    if config.get(ConfigKeys.OUTPUT_DIRECTORY) is not None:
         output_directory = (
-            utils.windows_to_wsl(config["output_directory"])
+            utils.windows_to_wsl(config[ConfigKeys.OUTPUT_DIRECTORY])
             if args.wsl
-            else config["output_directory"]
+            else config[ConfigKeys.OUTPUT_DIRECTORY]
         )
 
     if args.output is not None:
@@ -174,41 +183,6 @@ def parse_output_directory(args: argparse.Namespace, config: dict) -> str:
     return output_directory
 
 
-def parse_quant_dir(args: argparse.Namespace, config: dict) -> str:
-    """Parse custom quant path.
-    1. Use custom quant path from config file if specified.
-    2. Use custom quant path from command line if specified.
-
-    Parameters
-    ----------
-
-    args : argparse.Namespace
-        Command line arguments.
-
-    config : dict
-        Config dictionary.
-
-    Returns
-    -------
-
-    quant_dir : str
-        path to quant directory.
-    """
-
-    quant_dir = None
-    if "quant_dir" in config:
-        quant_dir = (
-            utils.windows_to_wsl(config["quant_dir"])
-            if args.wsl
-            else config["quant_dir"]
-        )
-
-    if args.quant_dir is not None:
-        quant_dir = utils.windows_to_wsl(args.quant_dir) if args.wsl else args.quant_dir
-
-    return quant_dir
-
-
 def parse_raw_path_list(args: argparse.Namespace, config: dict) -> list:
     """Parse raw file list.
     1. Use raw file list from config file if specified.
@@ -229,7 +203,7 @@ def parse_raw_path_list(args: argparse.Namespace, config: dict) -> list:
     raw_path_list : list
         List of raw files.
     """
-    config_raw_path_list = config.get("raw_path_list", [])
+    config_raw_path_list = config.get("raw_paths", [])
     raw_path_list = (
         utils.windows_to_wsl(config_raw_path_list) if args.wsl else config_raw_path_list
     )
@@ -259,71 +233,6 @@ def parse_raw_path_list(args: argparse.Namespace, config: dict) -> list:
     return raw_path_list
 
 
-def parse_library(args: argparse.Namespace, config: dict) -> str:
-    """Parse spectral library.
-    1. Use spectral library from config file if specified.
-    2. Use spectral library from command line if specified.
-
-    Parameters
-    ----------
-
-    args : argparse.Namespace
-        Command line arguments.
-
-    config : dict
-        Config dictionary.
-
-    Returns
-    -------
-
-    library : str
-        Spectral library.
-    """
-
-    library = None
-    if "library" in config:
-        library = (
-            utils.windows_to_wsl(config["library"]) if args.wsl else config["library"]
-        )
-
-    if args.library is not None:
-        library = utils.windows_to_wsl(args.library) if args.wsl else args.library
-
-    return library
-
-
-def parse_fasta(args: argparse.Namespace, config: dict) -> list:
-    """Parse fasta file list.
-    1. Use fasta file list from config file if specified.
-    2. Use fasta file list from command line if specified.
-
-    Parameters
-    ----------
-
-    args : argparse.Namespace
-        Command line arguments.
-
-    config : dict
-        Config dictionary.
-
-    Returns
-    -------
-
-    fasta_path_list : list
-        List of fasta files.
-    """
-
-    config_fasta_path_list = config.get("fasta_list", [])
-    fasta_path_list = (
-        utils.windows_to_wsl(config_fasta_path_list)
-        if args.wsl
-        else config_fasta_path_list
-    )
-    fasta_path_list += utils.windows_to_wsl(args.fasta) if args.wsl else args.fasta
-
-    return fasta_path_list
-
-
 def run(*args, **kwargs):
     # parse command line arguments
     args, unknown = parser.parse_known_args()
@@ -344,29 +253,30 @@ def run(*args, **kwargs):
         # print help message if no output directory specified
         parser.print_help()
 
-        print("No output directory specified.")
+        print("No output directory specified. Please do so via CL-argument or config.")
         return
     reporting.init_logging(output_directory)
 
-    quant_dir = parse_quant_dir(args, config)
+    # TODO revisit the multiple sources of raw files (cli, config, regex, ...)
     raw_path_list = parse_raw_path_list(args, config)
-    library_path = parse_library(args, config)
-    fasta_path_list = parse_fasta(args, config)
+    cli_params_config = {
+        **({ConfigKeys.RAW_PATHS: raw_path_list} if raw_path_list else {}),
+        **({ConfigKeys.LIBRARY_PATH: args.library} if args.library is not None else {}),
+        **({ConfigKeys.FASTA_PATHS: args.library} if args.fasta else {}),
+        **(
+            {ConfigKeys.QUANT_DIRECTORY: args.library}
+            if args.quant_dir is not None
+            else {}
+        ),
+    }
 
     # TODO rename all output_directory, output_folder => output_path, quant_dir->quant_path (except cli parameter)
 
     # important to suppress matplotlib output
     matplotlib.use("Agg")
 
     try:
-        SearchPlan(
-            output_directory,
-            raw_path_list,
-            library_path,
-            fasta_path_list,
-            config,
-            quant_dir,
-        ).run_plan()
+        SearchPlan(output_directory, config, cli_params_config).run_plan()
 
     except Exception as e:
         if isinstance(e, CustomError):
@@ -382,5 +292,5 @@ def run(*args, **kwargs):
 
 
 # uncomment for debugging:
-# if __name__ == "__main__":
-#     run()
+if __name__ == "__main__":
+    run()

alphadia/constants/default.yaml

@@ -1,10 +1,12 @@
 # configuration for the extraction plan
 version: 1
 
-library: null
+# These values are typically filled via CLI parameters
 output_directory: null
-raw_path_list: []
-output: null
+library_path: null
+raw_paths: []
+fasta_paths: []
+quant_directory: null
 
 general:
   thread_count: 10
@@ -63,31 +65,31 @@ library_prediction:
 #    composition: H(-2)2H(8)13C(2)
 custom_modifications:
   # Dimethyl @K channel decoy
-  Dimethyl:d12@K:
+  - name: Dimethyl:d12@K
     composition: H(-2)2H(8)13C(2)
 
   # Dimethyl @Any_N-term channel decoy
-  Dimethyl:d12@Any_N-term:
+  - name: Dimethyl:d12@Any_N-term
     composition: H(-2)2H(8)13C(2)
 
   # Dimethyl @Protein_N-term channel decoy
-  Dimethyl:d12@Protein_N-term:
+  - name: Dimethyl:d12@Protein_N-term
     composition: H(-2)2H(8)13C(2)
 
   # mTRAQ @K channel decoy
-  mTRAQ:d12@K:
+  - name: mTRAQ:d12@K
     composition: H(12)C(1)13C(10)15N(2)O(1)
 
   # mTRAQ @Any_N-term channel decoy
-  mTRAQ:d12@Any_N-term:
+  - name: mTRAQ:d12@Any_N-term
     composition: H(12)C(1)13C(14)15N(2)O(1)
 
   # mTRAQ @Protein_N-term channel decoy
-  mTRAQ:d12@Protein_N-term:
+  - name: mTRAQ:d12@Protein_N-term
     composition: H(12)C(1)13C(14)15N(2)O(1)
 
   # SILAC heavy @K channel decoy
-  Label:13C(12)@K:
+  - name: Label:13C(12)@K
     composition: C(12)
 
 search:

alphadia/constants/keys.py

@@ -22,3 +22,13 @@ class OutputKeys(metaclass=ConstantsClass):
     MS2_ERROR = "ms2_error"
     RT_ERROR = "rt_error"
     MOBILITY_ERROR = "mobility_error"
+
+
+class ConfigKeys(metaclass=ConstantsClass):
+    """String constants for accessing the config."""
+
+    OUTPUT_DIRECTORY = "output_directory"
+    LIBRARY_PATH = "library_path"
+    RAW_PATHS = "raw_paths"
+    FASTA_PATHS = "fasta_paths"
+    QUANT_DIRECTORY = "quant_directory"

alphadia/constants/multistep.yaml

@@ -12,6 +12,7 @@ transfer:
     enabled: True
 
   # override settings that could have been set by the user:
+  quant_directory: null
   general:
     save_library: False
     reuse_quant: False
@@ -23,6 +24,7 @@ library:
 #    predict: True
 
   # override settings that could have been set by the user:
+  quant_directory: null
   general:
     save_library: False
     reuse_quant: False
@@ -37,6 +39,7 @@ mbr:
   search:
     target_num_candidates: 5
   # override settings that could have been set by the user:
+  quant_directory: null
   general:
     reuse_quant: False
   library_prediction: