uniprotParser/index.js

"use strict";
// https://ftp.uniprot.org/pub/databases/uniprot/current_release/knowledgebase/taxonomic_divisions/uniprot_sprot_human.xml.gz
var __awaiter = (this && this.__awaiter) || function (thisArg, _arguments, P, generator) {
    function adopt(value) { return value instanceof P ? value : new P(function (resolve) { resolve(value); }); }
    return new (P || (P = Promise))(function (resolve, reject) {
        function fulfilled(value) { try { step(generator.next(value)); } catch (e) { reject(e); } }
        function rejected(value) { try { step(generator["throw"](value)); } catch (e) { reject(e); } }
        function step(result) { result.done ? resolve(result.value) : adopt(result.value).then(fulfilled, rejected); }
        step((generator = generator.apply(thisArg, _arguments || [])).next());
    });
};
var __asyncValues = (this && this.__asyncValues) || function (o) {
    if (!Symbol.asyncIterator) throw new TypeError("Symbol.asyncIterator is not defined.");
    var m = o[Symbol.asyncIterator], i;
    return m ? m.call(o) : (o = typeof __values === "function" ? __values(o) : o[Symbol.iterator](), i = {}, verb("next"), verb("throw"), verb("return"), i[Symbol.asyncIterator] = function () { return this; }, i);
    function verb(n) { i[n] = o[n] && function (v) { return new Promise(function (resolve, reject) { v = o[n](v), settle(resolve, reject, v.done, v.value); }); }; }
    function settle(resolve, reject, d, v) { Promise.resolve(v).then(function(v) { resolve({ value: v, done: d }); }, reject); }
};
var __importDefault = (this && this.__importDefault) || function (mod) {
    return (mod && mod.__esModule) ? mod : { "default": mod };
};
Object.defineProperty(exports, "__esModule", { value: true });
exports.getInteractionsFromEntry = exports.getEntryFromGeneName = exports.getEnrty = exports.readOffset = exports.makeIndex = void 0;
const fs_1 = __importDefault(require("fs"));
const readline_1 = __importDefault(require("readline"));
const fast_xml_parser_1 = require("fast-xml-parser");
const jsonata_1 = __importDefault(require("jsonata"));
const line$ = (path) => readline_1.default.createInterface({
    input: fs_1.default.createReadStream(path),
    crlfDelay: Infinity
});
const makeIndex = (filePath, indexPath) => __awaiter(void 0, void 0, void 0, function* () {
    var e_1, _a;
    indexPath = indexPath || filePath + '.jsi';
    let byteAcc = 0;
    const fromSel = new RegExp("^<entry");
    const toSel = new RegExp("^</entry>");
    const valSel = new RegExp('<accession>');
    let tmp = { values: [] };
    try {
        for (var _b = __asyncValues(line$(filePath)), _c; _c = yield _b.next(), !_c.done;) {
            const line = _c.value;
            if (fromSel.test(line))
                tmp['from'] = byteAcc;
            byteAcc += (line.length + 1);
            if (valSel.test(line))
                tmp['values'].push(line.match("<accession>(.*?)</accession>")[1]); // 'uck 
            if (toSel.test(line)) {
                yield fs_1.default.promises.appendFile(indexPath, tmp.values.join(';') + '\t' + tmp.from + '\t' + byteAcc + '\n');
                tmp = { values: [] };
            }
        }
    }
    catch (e_1_1) { e_1 = { error: e_1_1 }; }
    finally {
        try {
            if (_c && !_c.done && (_a = _b.return)) yield _a.call(_b);
        }
        finally { if (e_1) throw e_1.error; }
    }
});
exports.makeIndex = makeIndex;
const readOffset = (path, from, to) => {
    return new Promise((resolve, reject) => __awaiter(void 0, void 0, void 0, function* () {
        const size = to - from;
        const buffer = Buffer.alloc(size);
        let filehandle = null;
        try {
            filehandle = yield fs_1.default.promises.open(path, 'r+');
            yield filehandle.read(buffer, 0, buffer.length, from);
        }
        finally {
            if (filehandle) {
                yield filehandle.close();
                resolve(buffer.toString());
            }
        }
    }));
};
exports.readOffset = readOffset;
const getEntryOffset = (dbPath, accession) => __awaiter(void 0, void 0, void 0, function* () {
    var e_2, _d;
    const indexPath = dbPath + '.jsi';
    if (!fs_1.default.existsSync(indexPath))
        yield makeIndex(dbPath);
    const lineSel = new RegExp(accession);
    try {
        for (var _e = __asyncValues(line$(indexPath)), _f; _f = yield _e.next(), !_f.done;) {
            const line = _f.value;
            if (lineSel.test(line))
                return [Number(line.split('\t')[1]), Number(line.split('\t')[2])];
        }
    }
    catch (e_2_1) { e_2 = { error: e_2_1 }; }
    finally {
        try {
            if (_f && !_f.done && (_d = _e.return)) yield _d.call(_e);
        }
        finally { if (e_2) throw e_2.error; }
    }
    return [0, 0];
});
const getEnrty = (dbPath, accession) => __awaiter(void 0, void 0, void 0, function* () {
    const parser = new fast_xml_parser_1.XMLParser({
        ignoreAttributes: false,
        alwaysCreateTextNode: false,
        attributeNamePrefix: "",
        textNodeName: "value",
        allowBooleanAttributes: true,
    });
    const offsets = yield getEntryOffset(dbPath, accession);
    return parser.parse(yield readOffset(dbPath, offsets[0], offsets[1]));
});
exports.getEnrty = getEnrty;
const getEntryFromGeneName = (idmappingPath, dbPath, geneName) => __awaiter(void 0, void 0, void 0, function* () {
    var e_3, _g;
    const sel = new RegExp('Gene_Name\t' + geneName);
    let accessions = [];
    try {
        for (var _h = __asyncValues(line$(idmappingPath)), _j; _j = yield _h.next(), !_j.done;) {
            const line = _j.value;
            if (sel.test(line))
                accessions.push(line.split('\t')[0]);
        }
    }
    catch (e_3_1) { e_3 = { error: e_3_1 }; }
    finally {
        try {
            if (_j && !_j.done && (_g = _h.return)) yield _g.call(_h);
        }
        finally { if (e_3) throw e_3.error; }
    }
    return yield getEnrty(dbPath, accessions[0]);
});
exports.getEntryFromGeneName = getEntryFromGeneName;
const getInteractionsFromEntry = (json) => __awaiter(void 0, void 0, void 0, function* () {
    const uniprotIDs = Array.isArray(json.entry.accession) ? json.entry.accession : [json.entry.accession];
    // Comment interactant
    const res_inter = (0, jsonata_1.default)(`entry.comment[type="interaction"].interactant`).evaluate(json);
    let genes_interactant = [];
    if (res_inter) {
        genes_interactant = [...new Set(res_inter
                .filter((e) => !uniprotIDs.includes(e.label))
                .map((e) => e.label)
                .filter((e) => e))];
    }
    // Reference scope = INTERACTION WITH
    const scope_inter = (0, jsonata_1.default)(`entry.reference[scope ~> /INTERACTION WITH/i ]`).evaluate(json);
    let genes_scope_inter = [];
    if (scope_inter) {
        const comment_scope_inters = Array.isArray(scope_inter) ? scope_inter : [scope_inter];
        const comment_scope_inters_genes = comment_scope_inters.map((e) => (Object.assign({ txt: Array.isArray(e.scope) ? e.scope.filter((ee) => ee.match(/INTERACTION\ WITH/)).join() : e.scope }, e))).map((e) => (Object.assign({ interaction: e.txt.substring(e.txt.indexOf("INTERACTION WITH ") + "INTERACTION WITH ".length, e.txt.length) }, e)));
        genes_scope_inter = [...new Set(comment_scope_inters_genes.map((e) => e.interaction))].flatMap((e) => e.split(/; | AND /));
    }
    // Comment subunit
    const comment_subunit = (0, jsonata_1.default)(`entry.comment[type="subunit"].text.value`).evaluate(json);
    let comment_subunits_genes = [];
    if (comment_subunit) {
        const comment_subunits = Array.isArray(comment_subunit) ? comment_subunit : [comment_subunit];
        comment_subunits_genes = comment_subunits
            .flatMap((e) => e.replace(/ *\([^)]*\) */g, '').split(/\n/))
            .filter((e) => /Interacts/.test(e))
            .flatMap((e) => e.match(/[A-Z][A-Z0-9]{2,}/g));
    }
    let res = [...new Set([...genes_scope_inter, ...genes_interactant, ...comment_subunits_genes])].sort().filter((e) => typeof e === 'string').filter(_ => _);
    console.log(res);
    if (res.length > 0)
        res = res.flatMap((e) => e.match(/[A-Z][A-Z0-9]{2,}/g)).filter(_ => _);
    return res;
});
exports.getInteractionsFromEntry = getInteractionsFromEntry;