molstar
Version:
A comprehensive macromolecular library.
81 lines (80 loc) • 3.61 kB
JavaScript
;
Object.defineProperty(exports, "__esModule", { value: true });
exports.AlphaFoldPAEExample = void 0;
const jsx_runtime_1 = require("react/jsx-runtime");
/**
* Copyright (c) 2024 mol* contributors, licensed under MIT, See LICENSE file for more info.
*
* @author David Sehnal <david.sehnal@gmail.com>
*/
const client_1 = require("react-dom/client");
const app_1 = require("../../apps/viewer/app");
const ui_1 = require("../../extensions/model-archive/quality-assessment/pairwise/ui");
require("./index.html");
require("../../mol-plugin-ui/skin/light.scss");
class AlphaFoldPAEExample {
async init(options) {
this.plotContainerId = options.plotContainerId;
this.viewer = await app_1.Viewer.create(options.pluginContainerId, {
layoutIsExpanded: false,
layoutShowControls: false,
layoutShowLeftPanel: false,
layoutShowLog: false,
});
return this;
}
async load(afId) {
var _a, _b;
const id = afId.trim().toUpperCase();
const plotRoot = (0, client_1.createRoot)(document.getElementById(this.plotContainerId));
plotRoot.render((0, jsx_runtime_1.jsx)("div", { children: "Loading..." }));
await this.viewer.plugin.clear();
await this.viewer.loadAlphaFoldDb(id);
try {
const req = await fetch(`https://alphafold.ebi.ac.uk/files/AF-${id}-F1-predicted_aligned_error_v4.json`);
const json = await req.json();
const model = (_b = (_a = this.viewer.plugin.managers.structure.hierarchy.current.models[0]) === null || _a === void 0 ? void 0 : _a.cell.obj) === null || _b === void 0 ? void 0 : _b.data;
const metric = pairwiseMetricFromAlphaFoldDbJson(model, json);
plotRoot.render((0, jsx_runtime_1.jsx)("div", { className: 'msp-plugin', style: { background: 'white' }, children: (0, jsx_runtime_1.jsx)(ui_1.MAPairwiseScorePlot, { plugin: this.viewer.plugin, pairwiseMetric: metric, model: model }) }));
}
catch (err) {
plotRoot.render((0, jsx_runtime_1.jsxs)("div", { children: ["Error: ", String(err)] }));
}
}
}
exports.AlphaFoldPAEExample = AlphaFoldPAEExample;
function pairwiseMetricFromAlphaFoldDbJson(model, data) {
var _a;
if (!Array.isArray(data) || !((_a = data[0]) === null || _a === void 0 ? void 0 : _a.predicted_aligned_error))
return undefined;
const { residues, residueAtomSegments, atomSourceIndex } = model.atomicHierarchy;
const sortedResidueIndices = new Array(residues._rowCount).fill(0).map((_, i) => i);
sortedResidueIndices.sort((a, b) => {
const idxA = atomSourceIndex.value(residueAtomSegments.offsets[a]);
const idxB = atomSourceIndex.value(residueAtomSegments.offsets[b]);
return idxA - idxB;
});
const metricData = data[0].predicted_aligned_error;
const metric = {
id: 0,
name: 'AlphaFold DB PAE',
residueRange: [0, (residues._rowCount - 1)],
valueRange: [0, data[0].max_predicted_aligned_error],
values: {}
};
for (let i = 0; i < metricData.length; i++) {
const rA = sortedResidueIndices[i];
if (typeof rA !== 'number')
continue;
const row = metricData[i];
const xs = (metric.values[rA] = {});
for (let j = 0; j < row.length; j++) {
const rB = sortedResidueIndices[j];
if (typeof rB !== 'number')
continue;
xs[rB] = row[j];
}
}
return metric;
}
window.AlphaFoldPAEExample = new AlphaFoldPAEExample();