fusion-function 0.2.1__py3-none-any.whl

fusion_function/reference_catalog.json

@@ -0,0 +1,82 @@
+{
+  "catalog_version": 1,
+  "references": [
+    {
+      "species": "homo_sapiens",
+      "assembly": "GRCh38",
+      "release": 116,
+      "build_revision": 1,
+      "filename": "homo_sapiens.GRCh38.ensembl-116.ff-v1.build-1.sqlite.gz",
+      "compression": "gzip",
+      "archive_bytes": 2245743354,
+      "archive_sha256": "65222b7637ecfcb1ecff9d6862eeb20da9ac2d471a49d8c05568bc3f978c5866",
+      "database_bytes": 2727596032,
+      "database_sha256": "d3ffd6b684598d392f3d4525bd7f368ed3563dff8534a2c80d8cb361350aa1d4",
+      "doi": "10.5281/zenodo.23111108",
+      "url": "https://zenodo.org/records/23111108/files/homo_sapiens.GRCh38.ensembl-116.ff-v1.build-1.sqlite.gz?download=1",
+      "metadata": {
+        "format_version": "1",
+        "preprocessing_version": "1",
+        "cds_mapping_version": "2",
+        "feature_annotation_version": "3",
+        "transcript_payload_codec": "zlib-json-v1",
+        "sequence_chunk_size": "1048576",
+        "sequence_codec": "zlib"
+      },
+      "data_source_notices": {
+        "Ensembl": {
+          "terms": "Unrestricted project-generated data; third-party constraints may apply.",
+          "url": "https://www.ensembl.org/info/about/legal/disclaimer.html"
+        },
+        "UniProt Consortium": {
+          "terms": "CC BY 4.0. Credit UniProt, link to the license and identify modifications.",
+          "url": "https://www.uniprot.org/help/license",
+          "license_url": "https://creativecommons.org/licenses/by/4.0/"
+        },
+        "InterPro Consortium": {
+          "terms": "Current InterPro downloads: CC0 1.0. Retain historical-source notices.",
+          "url": "https://interpro-documentation.readthedocs.io/en/latest/license.html",
+          "license_url": "https://creativecommons.org/publicdomain/zero/1.0/"
+        },
+        "PANTHER": {
+          "terms": "Classification release 14.1 and 17.0 READMEs carry GPL-2.0-or-later notices. Confirm terms for the imported release and derived classifications before redistribution.",
+          "url": "https://data.pantherdb.org/ftp/sequence_classifications/"
+        },
+        "Ensembl member annotations": {
+          "terms": "Member-source terms are not replaced by Ensembl or InterPro terms. PROSITE database terms are CC BY-NC-ND 4.0 with commercial licensing; SMART models require a license. Confirm terms for derived match annotations.",
+          "url": "https://prosite.expasy.org/prosite_license.html",
+          "smart_url": "https://smart.embl.de/about.cgi"
+        }
+      },
+      "source_metadata": {
+        "format_version": "1",
+        "release": "116",
+        "species": "homo_sapiens",
+        "core_database": "homo_sapiens_core_116_38",
+        "assembly": "GRCh38",
+        "created_utc": "2026-10-01T01:17:06.715364+00:00",
+        "sequence_chunk_size": "1048576",
+        "sequence_codec": "zlib",
+        "tables": "[\"meta\", \"coord_system\", \"seq_region\", \"gene\", \"transcript\", \"exon\", \"exon_transcript\", \"translation\", \"analysis\", \"protein_feature\", \"interpro\", \"xref\", \"external_db\", \"object_xref\"]",
+        "sequence_types": "[\"dna\", \"pep\"]",
+        "counts": "{\"meta\": 235, \"coord_system\": 9, \"seq_region\": 268640, \"gene\": 87735, \"transcript\": 672310, \"exon\": 1422168, \"exon_transcript\": 5252227, \"translation\": 384051, \"analysis\": 100, \"protein_feature\": 6690040, \"interpro\": 25300, \"xref\": 2751838, \"external_db\": 501, \"object_xref\": 7737725, \"dna_sequences\": 706, \"pep_sequences\": 382428}",
+        "preprocessing_version": "1",
+        "preprocessing_counts": "{\"ready\": 368066, \"noncoding\": 288242, \"error\": 16002, \"protein_features\": 6296276}",
+        "preprocessed_utc": "2026-10-02T17:30:44.172625+00:00",
+        "interpro_entries_sha256": "1d247a42670d4a2913cec9f2a58813135c736d65128c1078d9c5eac797ffef1c",
+        "interpro_entry_types_loaded": "true",
+        "interpro_historical_entries": "{\"IPR001423\": \"109.0\", \"IPR004734\": \"106.0\", \"IPR009133\": \"109.0\", \"IPR010307\": \"108.0\", \"IPR017198\": \"106.0\", \"IPR018798\": \"107.0\", \"IPR019321\": \"108.0\", \"IPR024812\": \"109.0\", \"IPR026516\": \"109.0\", \"IPR026570\": \"109.0\", \"IPR029065\": \"107.0\", \"IPR031101\": \"109.0\", \"IPR032922\": \"107.0\", \"IPR033489\": \"109.0\", \"IPR033509\": \"109.0\", \"IPR039163\": \"109.0\", \"IPR039681\": \"109.0\", \"IPR039715\": \"109.0\", \"IPR039747\": \"109.0\", \"IPR042012\": \"106.0\", \"IPR043040\": \"107.0\", \"IPR043041\": \"107.0\", \"IPR043042\": \"107.0\", \"IPR043504\": \"107.0\", \"IPR044613\": \"109.0\", \"IPR044986\": \"109.0\", \"IPR045150\": \"107.0\", \"IPR048456\": \"106.0\", \"IPR050058\": \"108.0\", \"IPR050079\": \"108.0\", \"IPR050082\": \"108.0\", \"IPR050089\": \"107.0\", \"IPR050115\": \"108.0\", \"IPR050123\": \"107.0\", \"IPR050149\": \"108.0\", \"IPR050174\": \"108.0\", \"IPR050203\": \"108.0\", \"IPR050223\": \"108.0\", \"IPR050224\": \"108.0\", \"IPR050247\": \"108.0\", \"IPR050332\": \"109.0\", \"IPR050369\": \"108.0\", \"IPR050374\": \"108.0\", \"IPR050382\": \"109.0\", \"IPR050383\": \"109.0\", \"IPR050514\": \"109.0\", \"IPR050518\": \"108.0\", \"IPR050527\": \"107.0\", \"IPR050544\": \"109.0\", \"IPR050569\": \"109.0\", \"IPR050628\": \"107.0\", \"IPR050650\": \"107.0\", \"IPR050657\": \"107.0\", \"IPR050741\": \"107.0\", \"IPR050778\": \"107.0\", \"IPR050802\": \"107.0\", \"IPR050803\": \"106.0\", \"IPR050888\": \"108.0\", \"IPR050904\": \"108.0\", \"IPR050912\": \"107.0\", \"IPR050915\": \"108.0\", \"IPR050937\": \"107.0\", \"IPR050951\": \"107.0\", \"IPR050958\": \"109.0\", \"IPR050969\": \"109.0\", \"IPR051012\": \"107.0\", \"IPR051022\": \"107.0\", \"IPR051027\": \"107.0\", \"IPR051029\": \"107.0\", \"IPR051061\": \"107.0\", \"IPR051064\": \"107.0\", \"IPR051110\": \"109.0\", \"IPR051170\": \"109.0\", \"IPR051179\": \"108.0\", \"IPR051205\": \"108.0\", \"IPR051232\": \"108.0\", \"IPR051237\": \"107.0\", \"IPR051266\": \"109.0\", \"IPR051270\": \"107.0\", \"IPR051288\": \"108.0\", \"IPR051334\": \"107.0\", \"IPR051387\": \"108.0\", \"IPR051412\": \"109.0\", \"IPR051425\": \"109.0\", \"IPR051441\": \"109.0\", \"IPR051458\": \"109.0\", \"IPR051478\": \"107.0\", \"IPR051550\": \"109.0\", \"IPR051569\": \"107.0\", \"IPR051630\": \"108.0\", \"IPR051637\": \"109.0\", \"IPR051647\": \"109.0\", \"IPR051694\": \"109.0\", \"IPR051728\": \"109.0\", \"IPR051742\": \"108.0\", \"IPR051827\": \"109.0\", \"IPR051969\": \"107.0\", \"IPR052083\": \"109.0\", \"IPR052097\": \"108.0\", \"IPR052130\": \"109.0\", \"IPR052145\": \"109.0\", \"IPR052273\": \"109.0\", \"IPR052301\": \"109.0\", \"IPR052346\": \"106.0\", \"IPR052376\": \"109.0\", \"IPR052392\": \"109.0\", \"IPR052504\": \"109.0\", \"IPR052555\": \"109.0\", \"IPR052579\": \"109.0\", \"IPR052660\": \"109.0\", \"IPR052696\": \"109.0\", \"IPR052881\": \"109.0\", \"IPR053019\": \"109.0\", \"IPR053099\": \"109.0\", \"IPR053374\": \"107.0\", \"IPR054602\": \"107.0\", \"IPR055047\": \"107.0\", \"IPR055074\": \"107.0\", \"IPR055351\": \"107.0\", \"IPR056552\": \"107.0\"}",
+        "interpro_historical_sources": "[{\"release\": \"109.0\", \"sha256\": \"00b86a43b14652470d1110133b26f91314d5bddf9f164da76c494d7b322752ec\", \"url\": \"https://ftp.ebi.ac.uk/pub/databases/interpro/releases/109.0/entry.list\"}, {\"release\": \"108.0\", \"sha256\": \"11ddc8e6624a3ddaf40d5528e5aaf0a87308bd1c58a3f7087208091ab5eefc29\", \"url\": \"https://ftp.ebi.ac.uk/pub/databases/interpro/releases/108.0/entry.list\"}, {\"release\": \"107.0\", \"sha256\": \"333770e997fc6b88a836650ac5c9475b8934b4beb52a5a084d7d2e10dcc4af05\", \"url\": \"https://ftp.ebi.ac.uk/pub/databases/interpro/releases/107.0/entry.list\"}, {\"release\": \"106.0\", \"sha256\": \"15d98e0805a05cdbe4935ab2086e65c6e204181ae87c0b06cc33a315eea4db4a\", \"url\": \"https://ftp.ebi.ac.uk/pub/databases/interpro/releases/106.0/entry.list\"}]",
+        "interpro_preserved_metadata_sha256": "1d247a42670d4a2913cec9f2a58813135c736d65128c1078d9c5eac797ffef1c",
+        "panther_classifications_source": "{\"sha256\": \"0635357eeae76cb6dce8ce498f20e7ad5e101fe5a81717003c97acb2ed009b32\", \"url\": \"https://data.pantherdb.org/ftp/sequence_classifications/19.0/PANTHER_Sequence_Classification_files/PTHR19.0_human\", \"version\": \"19.0\"}",
+        "feature_annotation_version": "3",
+        "uniprot_features_source": "{\"sha256\": \"3296aef04049bbf83a21f3335436c9f2d3a00f6186fc10e4e66854ac37b2e5ab\", \"url\": \"https://rest.uniprot.org/uniprotkb/stream?format=xml&compressed=true&query=organism_id%3A9606%20AND%20reviewed%3Atrue\"}",
+        "uniprot_counts": "{\"entries\": 20431, \"features\": 332987, \"matched_proteins\": 97684, \"reviewed_human_entries\": 20431, \"sequence_mismatches\": 19693}",
+        "cds_mapping_version": "2",
+        "preprocessing_errors": "{\"CDS/peptide length mismatch\": {\"count\": 14362, \"example_transcripts\": [\"ENST00000361390\", \"ENST00000361453\", \"ENST00000362079\"]}, \"Unsupported transcript coordinate system\": {\"count\": 1640, \"example_transcripts\": [\"LRG_1t1\", \"LRG_1000t1\", \"LRG_1001t1\"]}}",
+        "uniprot_lookup_fingerprint": "7f8c19064714799b35cca12a4320c26e40f21faef2ac0e9a3220ad005779ca33",
+        "transcript_payload_codec": "zlib-json-v1"
+      }
+    }
+  ]
+}

fusion_function/uniprot.py

@@ -0,0 +1,315 @@
+"""Reviewed human UniProt features, imported once during reference preparation.
+
+XML is streamed with the standard library. Isoforms are reconstructed only from
+explicit UniProt splice variants; canonical features are transferred only across
+unchanged sequence. Ensembl mappings additionally require an exact whole-protein
+sequence match. No gene-name matching, approximate alignment or runtime network
+access is used.
+"""
+
+from __future__ import annotations
+
+import json
+import sqlite3
+from collections import OrderedDict
+from collections.abc import Iterator
+from pathlib import Path
+from typing import TypedDict
+from xml.etree import ElementTree as ET
+
+from .ensembl import FeatureEvidence
+
+
+UNIPROT_HUMAN_URL = (
+    "https://rest.uniprot.org/uniprotkb/stream?format=xml&compressed=true"
+    "&query=organism_id%3A9606%20AND%20reviewed%3Atrue"
+)
+NS = "{http://uniprot.org/uniprot}"
+FEATURE_TYPES = {
+    "domain": "domain",
+    "active site": "active_site",
+    "binding site": "binding_site",
+    "site": "conserved_site",
+    "motif": "motif",
+    "short sequence motif": "motif",
+}
+
+
+class CuratedFeature(TypedDict):
+    feature_id: str
+    feature_type: str
+    start: int
+    end: int
+    description: str
+    uniprot_accession: str
+    uniprot_isoform: str
+    evidence: list[FeatureEvidence]
+
+
+class CuratedProtein(TypedDict):
+    sequence: str
+    features: list[CuratedFeature]
+    accessions: list[str]
+    protein_ids: list[str]
+
+
+def exact_interval(element: ET.Element) -> tuple[int, int] | None:
+    """Reject uncertain/fuzzy coordinates rather than inventing exact bounds."""
+    location = element.find(NS + "location")
+    if location is None:
+        return None
+    position = location.find(NS + "position")
+    nodes = (
+        [position, position]
+        if position is not None
+        else [location.find(NS + "begin"), location.find(NS + "end")]
+    )
+    if any(
+        node is None
+        or node.get("status") not in {None, "certain"}
+        or not (node.get("position") or "").isdigit()
+        for node in nodes
+    ):
+        return None
+    start, end = [int(node.attrib["position"]) for node in nodes if node is not None]
+    return (start, end) if 1 <= start <= end else None
+
+
+def entry_proteins(entry: ET.Element) -> Iterator[CuratedProtein]:
+    """Yield canonical/described isoforms with safely mapped functional features."""
+    if entry.get("dataset") != "Swiss-Prot" or not any(
+        ref.get("type") == "NCBI Taxonomy" and ref.get("id") == "9606"
+        for ref in entry.findall(f"{NS}organism/{NS}dbReference")
+    ):
+        return
+    accessions = [node.text for node in entry.findall(NS + "accession") if node.text]
+    canonical = "".join((entry.findtext(NS + "sequence") or "").split())
+    if not accessions or not canonical:
+        raise ValueError("Reviewed human UniProt entry lacks accession or sequence")
+    accession = accessions[0]
+    evidence: dict[str, FeatureEvidence] = {}
+    for node in entry.findall(NS + "evidence"):
+        item: FeatureEvidence = {"code": node.get("type", "")}
+        reference = node.find(f"{NS}source/{NS}dbReference")
+        if reference is not None:
+            item["source"] = reference.get("type", "")
+            item["id"] = reference.get("id", "")
+        evidence[node.get("key", "")] = item
+    features = entry.findall(NS + "feature")
+    variants = {node.get("id"): node for node in features if node.get("type") == "splice variant"}
+    isoforms = entry.findall(f"{NS}comment[@type='alternative products']/{NS}isoform")
+    # A canonical entry need not declare alternative products.
+    sequences: dict[str, tuple[str, list[tuple[int, int, int]]]] = {accession: (canonical, [])}
+    displayed = accession
+    for isoform in isoforms:
+        sequence_element = isoform.find(NS + "sequence")
+        ids = [node.text for node in isoform.findall(NS + "id") if node.text]
+        if sequence_element is None or not ids:
+            continue
+        edits: list[tuple[int, int, str]] = []
+        if sequence_element.get("type") == "displayed":
+            displayed = ids[0]
+        elif sequence_element.get("type") == "described":
+            refs = (sequence_element.get("ref") or "").split()
+            if not refs:
+                continue
+            for variant_id in refs:
+                variant = variants.get(variant_id)
+                interval = exact_interval(variant) if variant is not None else None
+                if variant is None or interval is None or interval[1] > len(canonical):
+                    break
+                start, end = interval
+                original = variant.findtext(NS + "original")
+                if original and original != canonical[start - 1 : end]:
+                    break
+                replacement = variant.findtext(NS + "variation") or ""
+                edits.append((start, end, replacement))
+            if len(edits) != len(refs):
+                continue
+        else:
+            # External/not-described isoform sequences cannot be reconstructed.
+            continue
+        edits.sort()
+        if any(left[1] >= right[0] for left, right in zip(edits, edits[1:])):
+            continue
+        parts: list[str] = []
+        cursor = 0
+        for start, end, replacement in edits:
+            parts.extend((canonical[cursor : start - 1], replacement))
+            cursor = end
+        parts.append(canonical[cursor:])
+        reconstructed = "".join(parts)
+        if not reconstructed:
+            continue
+        for isoform_id in ids:
+            sequences[isoform_id] = (reconstructed, [(a, b, len(c)) for a, b, c in edits])
+    if displayed != accession:
+        sequences.pop(accession)
+    for isoform_id, (sequence, edits_map) in sequences.items():
+        # The accession and displayed isoform are aliases for the same product.
+        aliases = {isoform_id, accession} if isoform_id in {accession, displayed} else {isoform_id}
+        protein_ids: list[str] = []
+        for ref in entry.findall(f"{NS}dbReference[@type='Ensembl']"):
+            molecule = ref.find(NS + "molecule")
+            if (molecule.get("id") if molecule is not None else accession) not in aliases:
+                continue
+            protein_ids.extend(
+                node.get("value", "").split(".")[0]
+                for node in ref.findall(f"{NS}property[@type='protein sequence ID']")
+            )
+        curated: list[CuratedFeature] = []
+        for index, node in enumerate(features, 1):
+            feature_type = FEATURE_TYPES.get(node.get("type", ""))
+            interval = exact_interval(node)
+            if feature_type is None or interval is None:
+                continue
+            start, end = interval
+            location = node.find(NS + "location")
+            target = location.get("sequence") if location is not None else None
+            if target:
+                if target not in aliases:
+                    continue
+            else:
+                # A feature touching edited residues is omitted, even when the
+                # replacement has equal length. Unchanged features can shift.
+                if any(start <= b and end >= a for a, b, _ in edits_map):
+                    continue
+                shift = sum(size - (b - a + 1) for a, b, size in edits_map if b < start)
+                start, end = start + shift, end + shift
+            if not 1 <= start <= end <= len(sequence):
+                continue
+            description = (
+                node.get("description")
+                or node.findtext(f"{NS}ligand/{NS}name")
+                or node.get("type", "")
+                or "Unnamed UniProt feature"
+            )
+            curated.append(
+                {
+                    "feature_id": node.get("id") or f"{accession}:{index}",
+                    "feature_type": feature_type,
+                    "start": start,
+                    "end": end,
+                    "description": description,
+                    "uniprot_accession": accession,
+                    "uniprot_isoform": displayed if isoform_id == accession else isoform_id,
+                    "evidence": [
+                        evidence[key]
+                        for key in (node.get("evidence") or "").split()
+                        if key in evidence
+                    ],
+                }
+            )
+        yield {
+            "sequence": sequence,
+            "features": curated,
+            "accessions": list(
+                aliases | (set(accessions) if isoform_id in {accession, displayed} else set())
+            ),
+            "protein_ids": protein_ids,
+        }
+
+
+def import_features(db: sqlite3.Connection, path: Path) -> dict[str, int]:
+    """Store sequence-verified feature lookups in the caller's atomic transaction.
+
+    Cross-references narrow candidates before reading peptides. Each distinct
+    peptide is decompressed at most once from a bounded chunk cache; unmatched
+    UniProt isoforms and changed Ensembl sequences are counted and omitted.
+    """
+    from .data import LOG, fetch_sequence, input_progress, sqlite_activity
+
+    candidates: dict[str, set[str]] = {}
+    for stable_id, sequence_id in db.execute(
+        "SELECT stable_id, sequence_id FROM sequences WHERE kind='pep'"
+    ):
+        candidates.setdefault(stable_id, set()).add(sequence_id)
+    xrefs: dict[str, set[str]] = {}
+    tables = {row[0] for row in db.execute("SELECT name FROM sqlite_master WHERE type='table'")}
+    if "ensembl_object_xref" in tables and {"xref_id", "dbprimary_acc"} <= {
+        row[1] for row in db.execute("PRAGMA table_info(ensembl_xref)")
+    }:
+        # Some tiny fixtures do not have a full object_xref table.
+        columns = {row[1] for row in db.execute("PRAGMA table_info(ensembl_object_xref)")}
+        if {"xref_id", "ensembl_id", "ensembl_object_type"} <= columns:
+            external_join = ""
+            external_filter = ""
+            if (
+                "ensembl_external_db" in tables
+                and "external_db_id"
+                in {row[1] for row in db.execute("PRAGMA table_info(ensembl_xref)")}
+                and {"external_db_id", "db_name"}
+                <= {row[1] for row in db.execute("PRAGMA table_info(ensembl_external_db)")}
+            ):
+                external_join = "JOIN ensembl_external_db d ON d.external_db_id=x.external_db_id"
+                external_filter = "AND LOWER(d.db_name) LIKE '%uniprot%'"
+            with sqlite_activity("Read UniProt protein cross-references"):
+                for accession, stable_id in db.execute(f"""
+                    SELECT x.dbprimary_acc, t.stable_id FROM ensembl_xref x
+                    JOIN ensembl_object_xref o USING (xref_id)
+                    JOIN ensembl_translation t ON t.translation_id=o.ensembl_id
+                    {external_join}
+                    WHERE o.ensembl_object_type='Translation' {external_filter}
+                """):
+                    if stable_id in candidates:
+                        xrefs.setdefault(accession, set()).add(stable_id)
+    db.execute("DROP TABLE IF EXISTS ff_uniprot_features")
+    db.execute(
+        "CREATE TABLE ff_uniprot_features (sequence_id TEXT NOT NULL, feature_id TEXT NOT NULL, "
+        "payload TEXT NOT NULL, PRIMARY KEY(sequence_id, feature_id)) WITHOUT ROWID"
+    )
+    counts = {
+        "entries": 0,
+        "reviewed_human_entries": 0,
+        "matched_proteins": 0,
+        "features": 0,
+        "sequence_mismatches": 0,
+    }
+    matched: set[str] = set()
+    cache: OrderedDict[tuple[str, str, int], bytes] = OrderedDict()
+    with path.open("rb") as source:
+        compressed = source.read(2) == b"\x1f\x8b"
+    with input_progress(path, "Import reviewed human UniProt", compressed=compressed) as (
+        stream,
+        bar,
+        update,
+    ):
+        parser = ET.iterparse(stream, events=("start", "end"))
+        _, root = next(parser)
+        if root.tag != NS + "uniprot":
+            raise ValueError("Expected UniProt XML")
+        for event, entry in parser:
+            if event != "end" or entry.tag != NS + "entry":
+                continue
+            proteins = entry_proteins(entry)
+            reviewed = False
+            for protein in proteins:
+                reviewed = True
+                stable_ids = set(protein["protein_ids"])
+                for accession in protein["accessions"]:
+                    stable_ids.update(xrefs.get(accession, ()))
+                for stable_id in stable_ids:
+                    for sequence_id in candidates.get(stable_id, ()):
+                        peptide = fetch_sequence(db, "pep", sequence_id, _chunk_cache=cache)
+                        if peptide != protein["sequence"]:
+                            counts["sequence_mismatches"] += 1
+                            continue
+                        matched.add(sequence_id)
+                        for feature in protein["features"]:
+                            payload = json.dumps(feature, separators=(",", ":"))
+                            db.execute(
+                                "INSERT OR IGNORE INTO ff_uniprot_features VALUES (?, ?, ?)",
+                                (sequence_id, feature["feature_id"], payload),
+                            )
+            counts["entries"] += 1
+            counts["reviewed_human_entries"] += int(reviewed)
+            update()
+            if counts["entries"] % 100 == 0:
+                bar.set_postfix(entries=counts["entries"], matched=len(matched), refresh=False)
+            root.clear()  # Avoid retaining the full human XML tree in memory.
+    if counts["reviewed_human_entries"] == 0:
+        raise ValueError("UniProt XML contains no reviewed human entries")
+    counts["matched_proteins"] = len(matched)
+    counts["features"] = db.execute("SELECT COUNT(*) FROM ff_uniprot_features").fetchone()[0]
+    LOG.info("UniProt import: %s", counts)
+    return counts

fusion_function-0.2.1.dist-info/METADATA

@@ -0,0 +1,98 @@
+Metadata-Version: 2.4
+Name: fusion-function
+Version: 0.2.1
+Summary: Fusion frame and protein-domain prediction using a local Ensembl reference
+License-File: LICENSE
+Author: Caralyn Reisle
+Author-email: creisle@bcgsc.ca
+Requires-Python: >=3.11,<4.0
+Classifier: Programming Language :: Python :: 3
+Classifier: Programming Language :: Python :: 3.11
+Classifier: Programming Language :: Python :: 3.12
+Classifier: Programming Language :: Python :: 3.13
+Classifier: Programming Language :: Python :: 3.14
+Classifier: Programming Language :: Python :: 3.15
+Requires-Dist: tqdm (>=4.67,<5)
+Description-Content-Type: text/markdown
+
+# Fusion-function
+
+![Coverage](https://raw.githubusercontent.com/creisle/fusion_function/badges/badges/main/coverage.svg)
+
+Predict fusion reading frames and retained, disrupted, or excluded functional protein features from human GRCh38 transcript breakpoints. Predictions include effects of splicing, translation initiation, and premature termination. Annotation uses a preprocessed local SQLite reference.
+
+## Quick start
+
+```bash
+pip install fusion-function
+fusion-function prepare-data
+```
+
+Preparation downloads a compatible prebuilt reference for the latest Ensembl release when available, otherwise builds it from source. The reference is stored in the default user cache; source builds can take an hour or longer.
+
+Use the annotate_fusion_domains function to get information about the status of various domains in the expected fusion product
+
+```python
+from fusion_function import ReferenceDatabase, annotate_fusion_domains
+
+with ReferenceDatabase() as ref:
+    # ex. BCR::ABL1
+    result = annotate_fusion_domains(
+        transcript1_id="ENST00000305877",  # BCR
+        transcript2_id="ENST00000318560",  # ABL1
+        breakpoint1="22:23290413",
+        breakpoint2="9:130854064",
+        gene1_terminus="N",
+        gene2_terminus="C",
+        reference=ref,
+    )
+```
+
+This will return an object with the following shape. See the [api](./docs/api.md) for details.
+
+```json
+{
+    "frame_status": "in_frame",
+    "domains": [
+        {
+            "transcript_id": "ENST00000305877",
+            "interpro_id": "IPR036481",
+            "name": "Bcr-Abl oncoprotein oligomerisation domain superfamily",
+            "domain_type": "homologous_superfamily",
+            "start": 1,
+            "end": 67,
+            "sources": ["SuperFamily"],
+            "feature_ids": ["SSF69036"],
+            "breakpoint_based_status": "included",
+            "breakpoint_retained_percent": 100.0,
+            "post_splicing_status": "preserved",
+            "post_translation_status": "preserved"
+        },
+        ...
+    ],
+    "translation_start": {"ENST00000305877": "native_start_retained"}
+}
+```
+
+Analysis uses the newest prepared local release by default.
+
+## Documentation
+
+- [API](docs/api.md)
+- [Reference configuration and preparation](docs/reference.md)
+- [Methodology](docs/methodology.md)
+- [Testing](docs/testing.md)
+
+## Important Limitations
+
+- This package uses the splicing model defined by [MAVIS](https://github.com/bcgsc/mavis), this is non-exhaustive. It assumes splice sites to be disrupted based on a breakpoint being within 2bp but there are many other ways to disrupt splicing that are difficult to predict computationally (ex. deep intronic). This package only covers the standard scenarios
+- Currently we only support Hg38
+- Only exact breakpoints are supported
+- Predicted structural consequences do not establish fusion expression, oncogenicity, pathogenicity, or clinical actionability.
+
+## Citation
+
+This package ports and extends selected fusion-annotation logic from [MAVIS](https://github.com/bcgsc/mavis). Please cite:
+
+Reisle C, Mungall KL, Choo C, et al. MAVIS: merging, annotation, validation, and illustration of structural variants. *Bioinformatics*. 2019;35(3):515–517. PMID:30016509.
+

fusion_function-0.2.1.dist-info/RECORD

@@ -0,0 +1,16 @@
+fusion_function/__init__.py,sha256=kGDFqDAJP8D_9M43wiPD_rtymDlENkvrH6VR0Uxg6vI,258
+fusion_function/__main__.py,sha256=k1ocEWawweo1qCJWNFAAvyxz3tcY13dzvCenHszij30,48
+fusion_function/cli.py,sha256=DqDcyYK0R4d88hDe8D-uLE2cHMks95rxl7exScrqbHo,767
+fusion_function/data.py,sha256=Tu230UT1SAhjhRIMNTDGme_X5xCuG8vdVevfrS-m4mk,121039
+fusion_function/ensembl.py,sha256=Rip1k7rQlDrsNVPiF4xQxIMjjInoIDeMqIg5GmM21Gc,3889
+fusion_function/fusion.py,sha256=9_KPXMkvpkdrA9coMVQqILYnrnWGkm8xNi2_AwBDxUM,40495
+fusion_function/interpro.py,sha256=TVQxzR3xO5y1dQb5-_kHmkRNTvFNjGwVZpxw5Bt1Hjc,731
+fusion_function/prebuilt.py,sha256=Lyrc0WMGwormvrEkr8eac3Fg2xRF2CJ366vJl3WDon4,23713
+fusion_function/reference.py,sha256=OOgLN_4MuEzczYamfJ5rbQv3bSBWFLocjE2V9jV5Jho,4407
+fusion_function/reference_catalog.json,sha256=8REBPh6B46K-Z9eWp-Ef-gLKsyTlEZ4ifFoXcwMwAaU,9260
+fusion_function/uniprot.py,sha256=izymwyMI6fog3mUc2vwWuhgs80fyH8BfEeg4hCy6On0,13862
+fusion_function-0.2.1.dist-info/METADATA,sha256=zTidauqLnSTCufH65Tcr6eR8w2TamVHdvb0_-y18Ln0,3853
+fusion_function-0.2.1.dist-info/WHEEL,sha256=L-WvLdSBvJWHXv5zrQxDvNRnIgoBzLUpcbuiw19KCuQ,88
+fusion_function-0.2.1.dist-info/entry_points.txt,sha256=-cTv1JByye6d8iVffWHKz6yNApXduTlch2JooZK8KFY,60
+fusion_function-0.2.1.dist-info/licenses/LICENSE,sha256=OXLcl0T2SZ8Pmy2_dmlvKuetivmyPd5m1q-Gyd-zaYY,35149
+fusion_function-0.2.1.dist-info/RECORD,,

fusion_function-0.2.1.dist-info/WHEEL

@@ -0,0 +1,4 @@
+Wheel-Version: 1.0
+Generator: poetry-core 2.5.0
+Root-Is-Purelib: true
+Tag: py3-none-any

fusion_function-0.2.1.dist-info/entry_points.txt

@@ -0,0 +1,3 @@
+[console_scripts]
+fusion-function=fusion_function.cli:main
+