protein-quest 0.3.1__tar.gz → 0.3.2__tar.gz

{protein_quest-0.3.1 → protein_quest-0.3.2}/PKG-INFO

@@ -1,6 +1,6 @@
 Metadata-Version: 2.4
 Name: protein_quest
-Version: 0.3.1
+Version: 0.3.2
 Summary: Search/retrieve/filter proteins and protein structures
 Project-URL: Homepage, https://github.com/haddocking/protein-quest
 Project-URL: Issues, https://github.com/haddocking/protein-quest/issues
@@ -59,9 +59,11 @@ graph TB;
     searchpdbe -->|pdb_ids|fetchpdbe[Retrieve PDBe]
     searchaf --> |uniprot_accessions|fetchad(Retrieve AlphaFold)
     searchemdb -. emdb_ids .->fetchemdb[Retrieve EMDB]
-    fetchpdbe -->|mmcif_files_with_uniprot_acc| chainfilter{Filter on chain of uniprot}
-    chainfilter --> |mmcif_files| residuefilter{Filter on chain length}
-    fetchad -->|pdb_files| confidencefilter{Filter out low confidence}
+    fetchpdbe -->|mmcif_files_with_uniprot_acc| chainfilter{{Filter on chain of uniprot}}
+    chainfilter --> |mmcif_files| residuefilter{{Filter on chain length}}
+    fetchad -->|pdb_files| confidencefilter{{Filter out low confidence}}
+    confidencefilter --> |mmcif_files| ssfilter{{Filter on secondary structure}}
+    residuefilter --> |mmcif_files| ssfilter
     classDef dashedBorder stroke-dasharray: 5 5;
     goterm:::dashedBorder
     taxonomy:::dashedBorder
@@ -175,6 +177,18 @@ protein-quest filter residue  \
     ./filtered-chains ./filtered
 ```
 
+### To filter on secondary structure
+
+To filter on structure being mostly alpha helices and have no beta sheets.
+
+```shell
+protein-quest filter secondary-structure \
+    --ratio-min-helix-residues 0.5 \
+    --ratio-max-sheet-residues 0.0 \
+    --write-stats filtered-ss/stats.csv \
+    ./filtered-chains ./filtered-ss
+```
+
 ### Search Taxonomy
 
 ```shell

{protein_quest-0.3.1 → protein_quest-0.3.2}/README.md

@@ -29,9 +29,11 @@ graph TB;
     searchpdbe -->|pdb_ids|fetchpdbe[Retrieve PDBe]
     searchaf --> |uniprot_accessions|fetchad(Retrieve AlphaFold)
     searchemdb -. emdb_ids .->fetchemdb[Retrieve EMDB]
-    fetchpdbe -->|mmcif_files_with_uniprot_acc| chainfilter{Filter on chain of uniprot}
-    chainfilter --> |mmcif_files| residuefilter{Filter on chain length}
-    fetchad -->|pdb_files| confidencefilter{Filter out low confidence}
+    fetchpdbe -->|mmcif_files_with_uniprot_acc| chainfilter{{Filter on chain of uniprot}}
+    chainfilter --> |mmcif_files| residuefilter{{Filter on chain length}}
+    fetchad -->|pdb_files| confidencefilter{{Filter out low confidence}}
+    confidencefilter --> |mmcif_files| ssfilter{{Filter on secondary structure}}
+    residuefilter --> |mmcif_files| ssfilter
     classDef dashedBorder stroke-dasharray: 5 5;
     goterm:::dashedBorder
     taxonomy:::dashedBorder
@@ -145,6 +147,18 @@ protein-quest filter residue  \
     ./filtered-chains ./filtered
 ```
 
+### To filter on secondary structure
+
+To filter on structure being mostly alpha helices and have no beta sheets.
+
+```shell
+protein-quest filter secondary-structure \
+    --ratio-min-helix-residues 0.5 \
+    --ratio-max-sheet-residues 0.0 \
+    --write-stats filtered-ss/stats.csv \
+    ./filtered-chains ./filtered-ss
+```
+
 ### Search Taxonomy
 
 ```shell

{protein_quest-0.3.1 → protein_quest-0.3.2}/docs/notebooks/alphafold.ipynb

@@ -301,7 +301,7 @@
    "metadata": {},
    "outputs": [],
    "source": [
-    "query = ConfidenceFilterQuery(confidence=80, min_threshold=100, max_threshold=1000)"
+    "query = ConfidenceFilterQuery(confidence=80, min_residues=100, max_residues=1000)"
    ]
   },
   {
@@ -318,7 +318,7 @@
   },
   {
    "cell_type": "code",
-   "execution_count": 17,
+   "execution_count": null,
    "id": "6a6f8e3f",
    "metadata": {},
    "outputs": [
@@ -338,7 +338,7 @@
    "source": [
     "list(\n",
     "    filter_files_on_confidence(\n",
-    "        input_files, ConfidenceFilterQuery(confidence=80, min_threshold=100, max_threshold=1000), output_dir\n",
+    "        input_files, ConfidenceFilterQuery(confidence=80, min_residues=100, max_residues=1000), output_dir\n",
     "    )\n",
     ")"
    ]

{protein_quest-0.3.1 → protein_quest-0.3.2}/src/protein_quest/__version__.py

@@ -1,2 +1,2 @@
-__version__ = "0.3.1"
+__version__ = "0.3.2"
 """The version of the package."""

{protein_quest-0.3.1 → protein_quest-0.3.2}/src/protein_quest/alphafold/confidence.py

@@ -7,7 +7,10 @@ from pathlib import Path
 
 import gemmi
 
+from protein_quest.converter import Percentage, PositiveInt, converter
 from protein_quest.pdbe.io import write_structure
+from protein_quest.ss import nr_of_residues_in_total
+from protein_quest.utils import CopyMethod, copyfile
 
 """
 Methods to filter AlphaFoldDB structures on confidence scores.
@@ -73,13 +76,25 @@ class ConfidenceFilterQuery:
     Parameters:
         confidence: The confidence threshold for filtering residues.
             Residues with a pLDDT (b-factor) above this value are considered high confidence.
-        min_threshold: The minimum number of high-confidence residues required to keep the structure.
-        max_threshold: The maximum number of high-confidence residues required to keep the structure.
+        min_residues: The minimum number of high-confidence residues required to keep the structure.
+        max_residues: The maximum number of high-confidence residues required to keep the structure.
     """
 
-    confidence: float
-    min_threshold: int
-    max_threshold: int
+    confidence: Percentage
+    min_residues: PositiveInt
+    max_residues: PositiveInt
+
+
+base_query_hook = converter.get_structure_hook(ConfidenceFilterQuery)
+
+
+@converter.register_structure_hook
+def confidence_filter_query_hook(val, _type) -> ConfidenceFilterQuery:
+    result: ConfidenceFilterQuery = base_query_hook(val, _type)
+    if result.min_residues > result.max_residues:
+        msg = f"min_residues {result.min_residues} cannot be larger than max_residues {result.max_residues}"
+        raise ValueError(msg)
+    return result
 
 
 @dataclass
@@ -93,17 +108,20 @@ class ConfidenceFilterResult:
     """
 
     input_file: str
-    count: int
+    count: PositiveInt
     filtered_file: Path | None = None
 
 
-def filter_file_on_residues(file: Path, query: ConfidenceFilterQuery, filtered_dir: Path) -> ConfidenceFilterResult:
+def filter_file_on_residues(
+    file: Path, query: ConfidenceFilterQuery, filtered_dir: Path, copy_method: CopyMethod = "copy"
+) -> ConfidenceFilterResult:
     """Filter a single AlphaFoldDB structure file (*.pdb[.gz], *.cif[.gz]) based on confidence.
 
     Args:
         file: The path to the PDB file to filter.
         query: The confidence filter query.
         filtered_dir: The directory to save the filtered PDB file.
+        copy_method: How to copy when no residues have to be removed.
 
     Returns:
         result with filtered_file property set to Path where filtered PDB file is saved.
@@ -112,19 +130,24 @@ def filter_file_on_residues(file: Path, query: ConfidenceFilterQuery, filtered_d
     structure = gemmi.read_structure(str(file))
     residues = set(find_high_confidence_residues(structure, query.confidence))
     count = len(residues)
-    if count < query.min_threshold or count > query.max_threshold:
+    if count < query.min_residues or count > query.max_residues:
         # Skip structure that is outside the min and max threshold
         # just return number of high confidence residues
         return ConfidenceFilterResult(
             input_file=file.name,
             count=count,
         )
+    total_residues = nr_of_residues_in_total(structure)
     filtered_file = filtered_dir / file.name
-    new_structure = filter_out_low_confidence_residues(
-        structure,
-        residues,
-    )
-    write_structure(new_structure, filtered_file)
+    if count == total_residues:
+        # if no residues have to be removed then copy instead of slower gemmi writing
+        copyfile(file, filtered_file, copy_method)
+    else:
+        new_structure = filter_out_low_confidence_residues(
+            structure,
+            residues,
+        )
+        write_structure(new_structure, filtered_file)
     return ConfidenceFilterResult(
         input_file=file.name,
         count=count,
@@ -133,7 +156,10 @@ def filter_file_on_residues(file: Path, query: ConfidenceFilterQuery, filtered_d
 
 
 def filter_files_on_confidence(
-    alphafold_pdb_files: list[Path], query: ConfidenceFilterQuery, filtered_dir: Path
+    alphafold_pdb_files: list[Path],
+    query: ConfidenceFilterQuery,
+    filtered_dir: Path,
+    copy_method: CopyMethod = "copy",
 ) -> Generator[ConfidenceFilterResult]:
     """Filter AlphaFoldDB structures based on confidence.
 
@@ -141,6 +167,7 @@ def filter_files_on_confidence(
         alphafold_pdb_files: List of mmcif/PDB files from AlphaFoldDB to filter.
         query: The confidence filter query containing the confidence thresholds.
         filtered_dir: Directory where the filtered mmcif/PDB files will be saved.
+        copy_method: How to copy when a direct copy is possible.
 
     Yields:
         For each mmcif/PDB files yields whether it was filtered or not,
@@ -150,4 +177,4 @@ def filter_files_on_confidence(
     # In ../filter.py:filter_files_on_residues() we filter on number of residues on a file level
     # here we filter on file level and inside file remove low confidence residues
     for pdb_file in alphafold_pdb_files:
-        yield filter_file_on_residues(pdb_file, query, filtered_dir)
+        yield filter_file_on_residues(pdb_file, query, filtered_dir, copy_method)

{protein_quest-0.3.1 → protein_quest-0.3.2}/src/protein_quest/alphafold/fetch.py

@@ -9,17 +9,15 @@ from typing import Literal, cast, get_args
 
 from aiohttp_retry import RetryClient
 from aiopath import AsyncPath
-from cattrs.preconf.orjson import make_converter
 from tqdm.asyncio import tqdm
 from yarl import URL
 
 from protein_quest.alphafold.entry_summary import EntrySummary
+from protein_quest.converter import converter
 from protein_quest.utils import friendly_session, retrieve_files, run_async
 
 logger = logging.getLogger(__name__)
-converter = make_converter()
-"""cattrs converter to read AlphaFold summary JSON document."""
-converter.register_structure_hook(URL, lambda v, _: URL(v))
+
 
 DownloadableFormat = Literal[
     "summary",

{protein_quest-0.3.1 → protein_quest-0.3.2}/src/protein_quest/cli.py

@@ -23,13 +23,16 @@ from protein_quest.__version__ import __version__
 from protein_quest.alphafold.confidence import ConfidenceFilterQuery, filter_files_on_confidence
 from protein_quest.alphafold.fetch import DownloadableFormat, downloadable_formats
 from protein_quest.alphafold.fetch import fetch_many as af_fetch
+from protein_quest.converter import converter
 from protein_quest.emdb import fetch as emdb_fetch
 from protein_quest.filters import filter_files_on_chain, filter_files_on_residues
 from protein_quest.go import Aspect, allowed_aspects, search_gene_ontology_term, write_go_terms_to_csv
 from protein_quest.pdbe import fetch as pdbe_fetch
 from protein_quest.pdbe.io import glob_structure_files, locate_structure_file
+from protein_quest.ss import SecondaryStructureFilterQuery, filter_files_on_secondary_structure
 from protein_quest.taxonomy import SearchField, _write_taxonomy_csv, search_fields, search_taxon
 from protein_quest.uniprot import PdbResult, Query, search4af, search4emdb, search4pdb, search4uniprot
+from protein_quest.utils import CopyMethod, copy_methods, copyfile
 
 logger = logging.getLogger(__name__)
 
@@ -282,6 +285,22 @@ def _add_retrieve_emdb_parser(subparsers: argparse._SubParsersAction):
     parser.add_argument("output_dir", type=Path, help="Directory to store downloaded EMDB volume files")
 
 
+def _add_copy_method_argument(parser: argparse.ArgumentParser):
+    """Add copy method argument to parser."""
+    default_copy_method = "symlink"
+    if os.name == "nt":
+        # On Windows you need developer mode or admin privileges to create symlinks
+        # so we default to copying files instead of symlinking
+        default_copy_method = "copy"
+    parser.add_argument(
+        "--copy-method",
+        type=str,
+        choices=copy_methods,
+        default=default_copy_method,
+        help="How to copy files when no changes are needed to output file.",
+    )
+
+
 def _add_filter_confidence_parser(subparsers: argparse._SubParsersAction):
     """Add filter confidence subcommand parser."""
     parser = subparsers.add_parser(
@@ -312,6 +331,7 @@ def _add_filter_confidence_parser(subparsers: argparse._SubParsersAction):
             In CSV format with `<input_file>,<residue_count>,<passed>,<output_file>` columns.
             Use `-` for stdout."""),
     )
+    _add_copy_method_argument(parser)
 
 
 def _add_filter_chain_parser(subparsers: argparse._SubParsersAction):
@@ -347,8 +367,11 @@ def _add_filter_chain_parser(subparsers: argparse._SubParsersAction):
     )
     parser.add_argument(
         "--scheduler-address",
-        help="Address of the Dask scheduler to connect to. If not provided, will create a local cluster.",
+        help=dedent("""Address of the Dask scheduler to connect to.
+            If not provided, will create a local cluster.
+            If set to `sequential` will run tasks sequentially."""),
     )
+    _add_copy_method_argument(parser)
 
 
 def _add_filter_residue_parser(subparsers: argparse._SubParsersAction):
@@ -371,6 +394,7 @@ def _add_filter_residue_parser(subparsers: argparse._SubParsersAction):
     )
     parser.add_argument("--min-residues", type=int, default=0, help="Min residues in chain A")
     parser.add_argument("--max-residues", type=int, default=10_000_000, help="Max residues in chain A")
+    _add_copy_method_argument(parser)
     parser.add_argument(
         "--write-stats",
         type=argparse.FileType("w", encoding="UTF-8"),
@@ -381,6 +405,43 @@ def _add_filter_residue_parser(subparsers: argparse._SubParsersAction):
     )
 
 
+def _add_filter_ss_parser(subparsers: argparse._SubParsersAction):
+    """Add filter secondary structure subcommand parser."""
+    parser = subparsers.add_parser(
+        "secondary-structure",
+        help="Filter PDB/mmCIF files by secondary structure",
+        description="Filter PDB/mmCIF files by secondary structure",
+        formatter_class=ArgumentDefaultsRichHelpFormatter,
+    )
+    parser.add_argument("input_dir", type=Path, help="Directory with PDB/mmCIF files (e.g., from 'filter chain')")
+    parser.add_argument(
+        "output_dir",
+        type=Path,
+        help=dedent("""\
+            Directory to write filtered PDB/mmCIF files. Files are copied without modification.
+        """),
+    )
+    parser.add_argument("--abs-min-helix-residues", type=int, help="Min residues in helices")
+    parser.add_argument("--abs-max-helix-residues", type=int, help="Max residues in helices")
+    parser.add_argument("--abs-min-sheet-residues", type=int, help="Min residues in sheets")
+    parser.add_argument("--abs-max-sheet-residues", type=int, help="Max residues in sheets")
+    parser.add_argument("--ratio-min-helix-residues", type=float, help="Min residues in helices (relative)")
+    parser.add_argument("--ratio-max-helix-residues", type=float, help="Max residues in helices (relative)")
+    parser.add_argument("--ratio-min-sheet-residues", type=float, help="Min residues in sheets (relative)")
+    parser.add_argument("--ratio-max-sheet-residues", type=float, help="Max residues in sheets (relative)")
+    _add_copy_method_argument(parser)
+    parser.add_argument(
+        "--write-stats",
+        type=argparse.FileType("w", encoding="UTF-8"),
+        help=dedent("""
+            Write filter statistics to file. In CSV format with columns:
+            `<input_file>,<nr_residues>,<nr_helix_residues>,<nr_sheet_residues>,
+            <helix_ratio>,<sheet_ratio>,<passed>,<output_file>`.
+            Use `-` for stdout.
+        """),
+    )
+
+
 def _add_search_subcommands(subparsers: argparse._SubParsersAction):
     """Add search command and its subcommands."""
     parser = subparsers.add_parser(
@@ -422,6 +483,7 @@ def _add_filter_subcommands(subparsers: argparse._SubParsersAction):
     _add_filter_confidence_parser(subsubparsers)
     _add_filter_chain_parser(subsubparsers)
     _add_filter_residue_parser(subsubparsers)
+    _add_filter_ss_parser(subsubparsers)
 
 
 def _add_mcp_command(subparsers: argparse._SubParsersAction):
@@ -620,21 +682,22 @@ def _handle_filter_confidence(args: argparse.Namespace):
     # to get rid of duplication
     input_dir = structure(args.input_dir, Path)
     output_dir = structure(args.output_dir, Path)
-    confidence_threshold = structure(args.confidence_threshold, float)
-    # TODO add min/max
-    min_residues = structure(args.min_residues, int)
-    max_residues = structure(args.max_residues, int)
+
+    confidence_threshold = args.confidence_threshold
+    min_residues = args.min_residues
+    max_residues = args.max_residues
     stats_file: TextIOWrapper | None = args.write_stats
+    copy_method: CopyMethod = structure(args.copy_method, CopyMethod)  # pyright: ignore[reportArgumentType]
 
     output_dir.mkdir(parents=True, exist_ok=True)
     input_files = sorted(glob_structure_files(input_dir))
     nr_input_files = len(input_files)
     rprint(f"Starting confidence filtering of {nr_input_files} mmcif/PDB files in {input_dir} directory.")
-    query = structure(
+    query = converter.structure(
         {
             "confidence": confidence_threshold,
-            "min_threshold": min_residues,
-            "max_threshold": max_residues,
+            "min_residues": min_residues,
+            "max_residues": max_residues,
         },
         ConfidenceFilterQuery,
     )
@@ -643,7 +706,11 @@ def _handle_filter_confidence(args: argparse.Namespace):
         writer.writerow(["input_file", "residue_count", "passed", "output_file"])
 
     passed_count = 0
-    for r in tqdm(filter_files_on_confidence(input_files, query, output_dir), total=len(input_files), unit="file"):
+    for r in tqdm(
+        filter_files_on_confidence(input_files, query, output_dir, copy_method=copy_method),
+        total=len(input_files),
+        unit="file",
+    ):
         if r.filtered_file:
             passed_count += 1
         if stats_file:
@@ -656,9 +723,10 @@ def _handle_filter_confidence(args: argparse.Namespace):
 
 def _handle_filter_chain(args):
     input_dir = args.input_dir
-    output_dir = args.output_dir
+    output_dir = structure(args.output_dir, Path)
     pdb_id2chain_mapping_file = args.chains
-    scheduler_address = args.scheduler_address
+    scheduler_address = structure(args.scheduler_address, str | None)  # pyright: ignore[reportArgumentType]
+    copy_method: CopyMethod = structure(args.copy_method, CopyMethod)  # pyright: ignore[reportArgumentType]
 
     # make sure files in input dir with entries in mapping file are the same
     # complain when files from mapping file are missing on disk
@@ -683,18 +751,25 @@ def _handle_filter_chain(args):
         rprint("[red]No valid structure files found. Exiting.")
         sys.exit(1)
 
-    results = filter_files_on_chain(file2chain, output_dir, scheduler_address=scheduler_address)
+    results = filter_files_on_chain(
+        file2chain, output_dir, scheduler_address=scheduler_address, copy_method=copy_method
+    )
 
     nr_written = len([r for r in results if r.passed])
 
     rprint(f"Wrote {nr_written} single-chain PDB/mmCIF files to {output_dir}.")
 
+    for result in results:
+        if result.discard_reason:
+            rprint(f"[red]Discarding {result.input_file} ({result.discard_reason})[/red]")
+
 
 def _handle_filter_residue(args):
     input_dir = structure(args.input_dir, Path)
     output_dir = structure(args.output_dir, Path)
     min_residues = structure(args.min_residues, int)
     max_residues = structure(args.max_residues, int)
+    copy_method: CopyMethod = structure(args.copy_method, CopyMethod)  # pyright: ignore[reportArgumentType]
     stats_file: TextIOWrapper | None = args.write_stats
 
     if stats_file:
@@ -705,7 +780,9 @@ def _handle_filter_residue(args):
     input_files = sorted(glob_structure_files(input_dir))
     nr_total = len(input_files)
     rprint(f"Filtering {nr_total} files in {input_dir} directory by number of residues in chain A.")
-    for r in filter_files_on_residues(input_files, output_dir, min_residues=min_residues, max_residues=max_residues):
+    for r in filter_files_on_residues(
+        input_files, output_dir, min_residues=min_residues, max_residues=max_residues, copy_method=copy_method
+    ):
         if stats_file:
             writer.writerow([r.input_file, r.residue_count, r.passed, r.output_file])
         if r.passed:
@@ -716,6 +793,68 @@ def _handle_filter_residue(args):
         rprint(f"Statistics written to {stats_file.name}")
 
 
+def _handle_filter_ss(args):
+    input_dir = structure(args.input_dir, Path)
+    output_dir = structure(args.output_dir, Path)
+    copy_method: CopyMethod = structure(args.copy_method, CopyMethod)  # pyright: ignore[reportArgumentType]
+    stats_file: TextIOWrapper | None = args.write_stats
+
+    raw_query = {
+        "abs_min_helix_residues": args.abs_min_helix_residues,
+        "abs_max_helix_residues": args.abs_max_helix_residues,
+        "abs_min_sheet_residues": args.abs_min_sheet_residues,
+        "abs_max_sheet_residues": args.abs_max_sheet_residues,
+        "ratio_min_helix_residues": args.ratio_min_helix_residues,
+        "ratio_max_helix_residues": args.ratio_max_helix_residues,
+        "ratio_min_sheet_residues": args.ratio_min_sheet_residues,
+        "ratio_max_sheet_residues": args.ratio_max_sheet_residues,
+    }
+    query = converter.structure(raw_query, SecondaryStructureFilterQuery)
+    input_files = sorted(glob_structure_files(input_dir))
+    nr_total = len(input_files)
+    output_dir.mkdir(parents=True, exist_ok=True)
+
+    if stats_file:
+        writer = csv.writer(stats_file)
+        writer.writerow(
+            [
+                "input_file",
+                "nr_residues",
+                "nr_helix_residues",
+                "nr_sheet_residues",
+                "helix_ratio",
+                "sheet_ratio",
+                "passed",
+                "output_file",
+            ]
+        )
+
+    rprint(f"Filtering {nr_total} files in {input_dir} directory by secondary structure.")
+    nr_passed = 0
+    for input_file, result in filter_files_on_secondary_structure(input_files, query=query):
+        output_file: Path | None = None
+        if result.passed:
+            output_file = output_dir / input_file.name
+            copyfile(input_file, output_file, copy_method)
+            nr_passed += 1
+        if stats_file:
+            writer.writerow(
+                [
+                    input_file,
+                    result.stats.nr_residues,
+                    result.stats.nr_helix_residues,
+                    result.stats.nr_sheet_residues,
+                    round(result.stats.helix_ratio, 3),
+                    round(result.stats.sheet_ratio, 3),
+                    result.passed,
+                    output_file,
+                ]
+            )
+    rprint(f"Wrote {nr_passed} files to {output_dir} directory.")
+    if stats_file:
+        rprint(f"Statistics written to {stats_file.name}")
+
+
 def _handle_mcp(args):
     if find_spec("fastmcp") is None:
         msg = "Unable to start MCP server, please install `protein-quest[mcp]`."
@@ -742,6 +881,7 @@ HANDLERS: dict[tuple[str, str | None], Callable] = {
     ("filter", "confidence"): _handle_filter_confidence,
     ("filter", "chain"): _handle_filter_chain,
     ("filter", "residue"): _handle_filter_residue,
+    ("filter", "secondary-structure"): _handle_filter_ss,
     ("mcp", None): _handle_mcp,
 }
 

protein_quest-0.3.2/src/protein_quest/converter.py

@@ -0,0 +1,45 @@
+"""Convert json or dict to Python objects."""
+
+from cattrs.preconf.orjson import make_converter
+from yarl import URL
+
+type Percentage = float
+"""Type alias for percentage values (0.0-100.0)."""
+type Ratio = float
+"""Type alias for ratio values (0.0-1.0)."""
+type PositiveInt = int
+"""Type alias for positive integer values (>= 0)."""
+
+converter = make_converter()
+"""cattrs converter to read JSON document or dict to Python objects."""
+converter.register_structure_hook(URL, lambda v, _: URL(v))
+
+
+@converter.register_structure_hook
+def percentage_hook(val, _) -> Percentage:
+    value = float(val)
+    """Cattrs hook to validate percentage values."""
+    if not 0.0 <= value <= 100.0:
+        msg = f"Value {value} is not a valid percentage (0.0-100.0)"
+        raise ValueError(msg)
+    return value
+
+
+@converter.register_structure_hook
+def ratio_hook(val, _) -> Ratio:
+    """Cattrs hook to validate ratio values."""
+    value = float(val)
+    if not 0.0 <= value <= 1.0:
+        msg = f"Value {value} is not a valid ratio (0.0-1.0)"
+        raise ValueError(msg)
+    return value
+
+
+@converter.register_structure_hook
+def positive_int_hook(val, _) -> PositiveInt:
+    """Cattrs hook to validate positive integer values."""
+    value = int(val)
+    if value < 0:
+        msg = f"Value {value} is not a valid positive integer (>= 0)"
+        raise ValueError(msg)
+    return value