Getting started¶

Requirements¶

Prozor supports Python 3.12 and newer.

Installation¶

Until the first package-index release, install directly from GitHub:

python -m pip install "prozor @ git+https://github.com/anndata-omics-bridge/prozor.git"

Both the accelerated ahocorasick_rs implementation and the portable ahocorapy fallback are installed. Public matching operations default to backend="auto": Rust is selected when importable, otherwise matching falls back to pure Python. Once released on a package index, install with pip install prozor.

Match peptides against proteins¶

Use a mapping when all proteins are already in memory:

from prozor.api import annotate_peptides

proteins = {
    "P1": "MYPEPTIDESEQUENCE",
    "P2": "XXSEQUENCEXX",
}
result = annotate_peptides(["PEPTIDE", "SEQUENCE"], proteins)

for match in result:
    print(match.peptide, match.protein_id, match.start, match.end)

The call above uses Rust by default. Pass backend="ahocorapy" to select the portable implementation explicitly.

Use a one-pass iterable when a consumer already streams FASTA or database records:

from collections.abc import Iterator
from dataclasses import dataclass

from prozor.api import ProteinSequenceRecord, annotate_peptides_streaming


@dataclass(frozen=True, slots=True)
class Protein:
    id: str
    sequence: str


def protein_records() -> Iterator[ProteinSequenceRecord]:
    yield Protein("P1", "MYPEPTIDESEQUENCE")
    yield Protein("P2", "XXSEQUENCEXX")


result = annotate_peptides_streaming(
    ["PEPTIDE", "SEQUENCE"],
    protein_records(),
)

Important

Prozor matches the strings it receives. It does not uppercase sequences, remove modifications, apply I/L equivalence, classify decoys, or parse FASTA headers. Those policies belong to the consuming application.

Infer protein groups¶

Convert occurrence-level matches to unique peptide--protein edges and run greedy parsimony. Repeated sites collapse naturally in the set:

from prozor.api import greedy_parsimony

edges = {(match.peptide, match.protein_id) for match in result}
inference = greedy_parsimony(edges)

for group in inference:
    print(group.protein_id, group.peptides)

See Protein inference for grouping, tie resolution, and subsumption semantics.