| 123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960616263646566676869707172737475767778798081828384858687888990919293949596979899100101102103104105106107108109110111112113114115116117118119120121122123124125126127128129130131132133134135136137138139140141142143144145146147148149150151152153154155156157158159160161162163164165166167168169170171172173174175176177178179180181182183 |
- "use strict";
- 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.getAllSymbols = exports.getN = exports.getGOterms = exports.getSymbol = void 0;
- const fs_1 = __importDefault(require("fs"));
- const readline_1 = __importDefault(require("readline"));
- const stats_1 = require("@stdlib/stats");
- const line$ = (path) => readline_1.default.createInterface({
- input: fs_1.default.createReadStream(path),
- crlfDelay: Infinity
- });
- // http://geneontology.org/docs/guide-go-evidence-codes/
- // http://wiki.geneontology.org/index.php/Category:Evidence_Codes
- // http://current.geneontology.org/ontology/go-basic.obo
- // http://current.geneontology.org/annotations/goa_human.gaf.gz
- const getSymbol = (symbol, goaPath, oboPath) => __awaiter(void 0, void 0, void 0, function* () {
- var e_1, _a;
- var _b;
- const header = [
- 'database', 'ID', 'Symbol', 'Qualifier',
- 'GO_Term', 'Evidence', 'Evidence_Code',
- 'With', 'From', 'Name', 'Alternative_symbols',
- 'Class', 'Taxon', 'Date', 'Origin'
- ];
- const tester = new RegExp('\t' + symbol + '\t');
- const separator = new RegExp('\\|');
- const results = [];
- try {
- for (var _c = __asyncValues(line$(goaPath)), _d; _d = yield _c.next(), !_d.done;) {
- const line = _d.value;
- if (tester.test(line))
- results.push(line.split('\t').filter((e) => e).reduce((p, c, i) => (Object.assign(Object.assign({}, p), { [header[i]]: separator.test(c) ? c.split('|') : c })), {}));
- }
- }
- catch (e_1_1) { e_1 = { error: e_1_1 }; }
- finally {
- try {
- if (_d && !_d.done && (_a = _c.return)) yield _a.call(_c);
- }
- finally { if (e_1) throw e_1.error; }
- }
- const subTerms = results.map(e => e === null || e === void 0 ? void 0 : e.GO_Term);
- const cacheTerms = yield getGOterms(subTerms, oboPath);
- for (let index = 0; index < results.length; index++) {
- const goTerm = (_b = results[index]) === null || _b === void 0 ? void 0 : _b.GO_Term;
- results[index]['GO_Term'] = cacheTerms.filter(e => e.id === goTerm)[0];
- }
- return results;
- });
- exports.getSymbol = getSymbol;
- const getGOterms = (terms, oboPath) => __awaiter(void 0, void 0, void 0, function* () {
- var e_2, _e;
- terms = Array.isArray(terms) ? terms : [terms];
- const testerList = terms.map(e => new RegExp('id: ' + e));
- let delim = false;
- const results = [];
- let result = {};
- try {
- for (var _f = __asyncValues(line$(oboPath)), _g; _g = yield _f.next(), !_g.done;) {
- const line = _g.value;
- if (testerList.some(rx => rx.test(line)))
- delim = true;
- if (line === '' && delim) {
- delim = false;
- results.push(result);
- result = {};
- }
- if (delim)
- result[line.split(': ')[0]] = line.split(': ')[1];
- }
- }
- catch (e_2_1) { e_2 = { error: e_2_1 }; }
- finally {
- try {
- if (_g && !_g.done && (_e = _f.return)) yield _e.call(_f);
- }
- finally { if (e_2) throw e_2.error; }
- }
- return results;
- });
- exports.getGOterms = getGOterms;
- const getN = (terms, qualifiers, goaPath, oboPath) => __awaiter(void 0, void 0, void 0, function* () {
- var e_3, _h;
- terms = Array.isArray(terms) ? terms : [terms];
- qualifiers = Array.isArray(qualifiers) ? qualifiers : [qualifiers];
- let qualifs = {};
- let termsList = [];
- for (const term of terms) {
- termsList.push(Object.assign(Object.assign({ nTotal: 0, genes: [] }, (yield getGOterms(term, oboPath))[0]), { test: new RegExp('\t' + term + '\t') }));
- }
- qualifiers.map((e) => qualifs[e] = {});
- try {
- for (var _j = __asyncValues(line$(goaPath)), _k; _k = yield _j.next(), !_k.done;) {
- const line = _k.value;
- const t = termsList.filter((rx) => rx.test.test(line));
- if (t.length > 0) {
- termsList = termsList.map((rx) => {
- if (rx.test.test(line)) {
- const genes = [...new Set([...(rx === null || rx === void 0 ? void 0 : rx.genes), line.split('\t')[2]])];
- return Object.assign(Object.assign({}, rx), { nTotal: genes.length, genes });
- }
- else {
- return rx;
- }
- });
- }
- }
- }
- catch (e_3_1) { e_3 = { error: e_3_1 }; }
- finally {
- try {
- if (_k && !_k.done && (_h = _j.return)) yield _h.call(_j);
- }
- finally { if (e_3) throw e_3.error; }
- }
- return termsList.map((e) => {
- delete e['test'];
- return e;
- });
- });
- exports.getN = getN;
- const headerGOA = [
- 'database', 'ID', 'Symbol', 'Qualifier',
- 'GO_Term', 'Evidence', 'Evidence_Code',
- 'With', 'From', 'Name', 'Alternative_symbols',
- 'Class', 'Taxon', 'Date', 'Origin'
- ];
- const getAllSymbols = (symbols, qualifier, goaPath, oboPath) => __awaiter(void 0, void 0, void 0, function* () {
- var e_4, _l;
- symbols = Array.isArray(symbols) ? symbols : [symbols];
- const testerList = [...new Set(symbols)].map(e => new RegExp('\t' + e + '\t' + qualifier + '\t'));
- const allGoa = [];
- const allGenes = {};
- try {
- for (var _m = __asyncValues(line$(goaPath)), _o; _o = yield _m.next(), !_o.done;) {
- const line = _o.value;
- if (testerList.some(rx => rx.test(line))) {
- allGoa.push(line.split('\t').
- reduce((p, c, i) => (Object.assign(Object.assign({}, p), { [headerGOA[i]]: c })), {}));
- }
- allGenes[line.split('\t')[2]] = {};
- }
- }
- catch (e_4_1) { e_4 = { error: e_4_1 }; }
- finally {
- try {
- if (_o && !_o.done && (_l = _m.return)) yield _l.call(_m);
- }
- finally { if (e_4) throw e_4.error; }
- }
- const nAllGenes = Object.keys(allGenes).length;
- console.log(nAllGenes);
- const allIDs = yield getN([...new Set(allGoa.map((e) => e.GO_Term))], qualifier, goaPath, oboPath);
- return allIDs.map((e) => {
- const observedGenes = symbols.filter((ee) => e.genes.includes(ee));
- // const observedGenesNotIn = symbols.filter((ee:string)=> !e.genes.includes(ee))
- // const allIn = Object.keys(allGenes).filter((ee:any)=>)
- const diff = e.genes.length - observedGenes.length;
- return Object.assign({ n: observedGenes.length, prop: observedGenes.length / e.genes.length, chi2test: (0, stats_1.chi2test)([[observedGenes.length,
- diff < 0 ? 0 : diff],
- [symbols.length,
- nAllGenes]]).pValue, observedGenes }, e);
- }).sort((a, b) => b.chi2test - a.chi2test);
- });
- exports.getAllSymbols = getAllSymbols;
|