AlphaPept Protein Group Reader

[lucas-diedrich]

Implements the protein group reader for AlphaPept protein group reports. This is the first reader that requires regex logic for reading.

Key challenges are that alphapept dataframes do not contain a name in the feature column and that feature names are extracted from custom fasta headers, i.e. it is difficult to parse the feature names.

• Implements a custom _pre_process logic that supports standard uniprot + ensembl entries.
• The parsing logic is implemented in a class-specific function _parse_alphapept_index, which is tested based on the examples provided in the example data.

[Copilot]

Pull Request Overview

This PR implements a new protein group reader for AlphaPept search engine outputs, enabling the parsing of AlphaPept protein group reports with custom regex-based protein identifier parsing.

• Adds AlphaPeptPGReader class with custom preprocessing logic for parsing protein identifiers from FASTA headers
• Extends the protein group column schema to support Ensembl IDs, source databases, and decoy indicators
• Implements comprehensive test coverage for the new reader functionality

Reviewed Changes

Copilot reviewed 9 out of 9 changed files in this pull request and generated 3 comments.

Show a summary per file

File	Description
alphabase/pg_reader/alphapept_pg_reader.py	Main implementation of AlphaPept reader with regex-based protein identifier parsing
alphabase/pg_reader/keys.py	Extends column schema with new fields for Ensembl IDs, source DB, and decoy indicators
alphabase/pg_reader/pg_reader.py	Modifies base reader to handle cases where no feature columns are specified
alphabase/constants/const_files/pg_reader.yaml	Configuration for AlphaPept reader with default measurement regex
alphabase/pg_reader/__init__.py	Exports the new AlphaPept reader class
tests/unit/pg_reader/test_alphapept_pg_reader.py	Comprehensive unit tests for protein identifier parsing logic
tests/integration/test_pg_readers.py	Integration tests for file import functionality
tests/integration/test_pg_reader_provider.py	Tests for reader provider integration
tests/integration/conftest.py	Test fixture for AlphaPept example data

[lucas-diedrich]

• Question: How to handle REV sequences

Currently, reverse (decoy) sequences are marked with a designated flag (is_decoy), and the respective database entries are marked with a REV__ prefix. However, IDs (uniprot IDs, ensembl IDs) and names are left unchanged. Is this desirable?

Example

# REV__sp|Q13085|ACACA_HUMAN
{"source_db": "REV__sp", "uniprot_ids": "Q13085", "ensembl_ids": "na", "proteins": "ACACA_HUMAN", "is_decoy": True}

[lucas-diedrich]

Added functionality to pre-configure regular expressions:

Configuration in pg_reader.yaml:

alphapept:
  ...
  measurement_regex:
    raw: "^.*(?<!_LFQ)$"
    lfq: "_LFQ$"

Usage

Get available pre-configured regular expressions:

AlphaPeptPGReader.get_preconfigured_regex()
 > {'raw': '^.*(?<!_LFQ)$', 'lfq': '_LFQ$'}

Read

reader = AlphaPeptPGReader(measurement_regex="lfq")
results = reader.import_file(...)
# reads LFQ values

[lucas-diedrich]

Added explicit equivalence test in integration testing.