Merge pull request #410 from MannLabs/change_config_handling

Change config handling

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=[],
 )
@@ -58,43 +64,46 @@
     "--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,
 )
 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,
 )
@@ -148,7 +157,7 @@ def _get_raw_path_list_from_args_and_config(
         list: a list of file paths that match the specified regex pattern.
     """
 
-    raw_path_list = config.get("raw_path_list", [])
+    raw_path_list = config.get(ConfigKeys.RAW_PATHS, [])
     raw_path_list += args.file
 
     if (config_directory := config.get("directory")) is not None:
@@ -172,17 +181,6 @@ def _get_raw_path_list_from_args_and_config(
     return raw_path_list
 
 
-def _get_fasta_list_from_args_and_config(
-    args: argparse.Namespace, config: dict
-) -> list:
-    """Parse fasta file list from command line arguments and config file, merging them if both are given."""
-
-    fasta_path_list = config.get("fasta_list", [])
-    fasta_path_list += args.fasta
-
-    return fasta_path_list
-
-
 def run(*args, **kwargs):
     # parse command line arguments
     args, unknown = parser.parse_known_args()
@@ -203,33 +201,31 @@ def run(*args, **kwargs):
     )
     if output_directory is None:
         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 = _get_from_args_or_config(
-        args, user_config, args_key="quant_dir", config_key="quant_dir"
-    )
-    raw_path_list = _get_raw_path_list_from_args_and_config(args, user_config)
-    library_path = _get_from_args_or_config(
-        args, user_config, args_key="library", config_key="library"
-    )
-    fasta_path_list = _get_fasta_list_from_args_and_config(args, user_config)
+    # TODO revisit the multiple sources of raw files (cli, config, regex, ...)
+    raw_paths = _get_raw_path_list_from_args_and_config(args, user_config)
+    cli_params_config = {
+        **({ConfigKeys.RAW_PATHS: raw_paths} if raw_paths 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=raw_path_list,
-            library_path=library_path,
-            fasta_path_list=fasta_path_list,
-            config=user_config,
-            quant_path=quant_dir,
-        ).run_plan()
+        SearchPlan(output_directory, user_config, cli_params_config).run_plan()
 
     except Exception as e:
         if isinstance(e, CustomError):

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 StatOutputKeys(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:

alphadia/exceptions.py

@@ -20,6 +20,9 @@ def msg(self):
     def detail_msg(self):
         return self._detail_msg
 
+    def __str__(self):
+        return f"{self._error_code}: {self._msg} {self._detail_msg}"
+
 
 class BusinessError(CustomError):
     """Custom error class for 'business' errors.
@@ -59,3 +62,33 @@ class NotDiaDataError(BusinessError):
     _error_code = "NOT_DIA_DATA"
 
     _msg = "Could not find cycle shape. Please check if this is a valid DIA data set."
+
+
+class ConfigError(BusinessError):
+    """Raise when something is wrong with the provided configuration."""
+
+    _error_code = "CONFIG_ERROR"
+
+    _msg = "Malformed configuration file(s)."
+    _key = ""
+    _config_name = ""
+
+    def __init__(self, key: str, config_name: str):
+        self._key = key
+        self._config_name = config_name
+
+
+class KeyAddedConfigError(ConfigError):
+    """Raise when a key should be added to a config."""
+
+    def __init__(self, key: str, config_name: str):
+        super().__init__(key, config_name)
+        self._detail_msg = f"Defining new keys is not allowed when updating a config: key='{self._key}', config_name='{self._config_name}'"
+
+
+class TypeMismatchConfigError(ConfigError):
+    """Raise when the type of a value does not match the default type."""
+
+    def __init__(self, key: str, config_name: str, extra_msg: str):
+        super().__init__(key, config_name)
+        self._detail_msg = f"Types of values must match default config: key='{self._key}', config_name='{self._config_name}', types='{extra_msg}'"

alphadia/libtransform.py

@@ -40,10 +40,10 @@ def __call__(self, *args: typing.Any) -> typing.Any:
             return self.forward(*args)
         else:
             logger.critical(
-                f"Input {input} failed validation for {self.__class__.__name__}"
+                f"Input {args} failed validation for {self.__class__.__name__}"
             )
             raise ValueError(
-                f"Input {input} failed validation for {self.__class__.__name__}"
+                f"Input {args} failed validation for {self.__class__.__name__}"
             )
 
     def validate(self, *args: typing.Any) -> bool: