From e9e27f32408d524820501c2a57ed3203ee1ed50d Mon Sep 17 00:00:00 2001
From: Alexander Rose <alexander.rose@weirdbyte.de>
Date: Thu, 28 Feb 2019 00:08:38 -0800
Subject: [PATCH] wip, basic gro format support
---
src/mol-io/reader/_spec/gro.spec.ts | 6 +-
src/mol-io/reader/cif/data-model.ts | 42 +++-
src/mol-io/reader/gro/parser.ts | 4 +-
src/mol-model-formats/structure/gro.ts | 236 +++++++-------------
src/mol-model-formats/structure/pdb.ts | 2 +-
src/mol-plugin/state/actions/data-format.ts | 3 +-
src/mol-plugin/state/actions/structure.ts | 42 +++-
src/mol-plugin/state/transforms/model.ts | 21 +-
8 files changed, 185 insertions(+), 171 deletions(-)
diff --git a/src/mol-io/reader/_spec/gro.spec.ts b/src/mol-io/reader/_spec/gro.spec.ts
index 055e4e61b..16a120a38 100644
--- a/src/mol-io/reader/_spec/gro.spec.ts
+++ b/src/mol-io/reader/_spec/gro.spec.ts
@@ -5,7 +5,7 @@
* @author David Sehnal <david.sehnal@gmail.com>
*/
-import Gro from '../gro/parser'
+import { parseGRO } from '../gro/parser'
const groString = `MD of 2 waters, t= 4.2
6
@@ -26,7 +26,7 @@ const groStringHighPrecision = `Generated by trjconv : 2168 system t= 15.00000
describe('gro reader', () => {
it('basic', async () => {
- const parsed = await Gro(groString).run();
+ const parsed = await parseGRO(groString).run();
if (parsed.isError) {
console.log(parsed)
@@ -57,7 +57,7 @@ describe('gro reader', () => {
});
it('high precision', async () => {
- const parsed = await Gro(groStringHighPrecision).run();
+ const parsed = await parseGRO(groStringHighPrecision).run();
if (parsed.isError) {
console.log(parsed)
diff --git a/src/mol-io/reader/cif/data-model.ts b/src/mol-io/reader/cif/data-model.ts
index 855cecb9b..2800437dc 100644
--- a/src/mol-io/reader/cif/data-model.ts
+++ b/src/mol-io/reader/cif/data-model.ts
@@ -132,7 +132,7 @@ export namespace CifField {
}
}
- export function ofNumbers(values: number[]): CifField {
+ export function ofNumbers(values: ArrayLike<number>): CifField {
const rowCount = values.length;
const str: CifField['str'] = row => { return '' + values[row]; };
const float: CifField['float'] = row => values[row];
@@ -196,6 +196,46 @@ export namespace CifField {
toFloatArray: params => ColumnHelpers.createAndFillArray(rowCount, float, params)
}
}
+
+ export function ofColumn(column: Column<any>): CifField {
+ const { rowCount, valueKind, areValuesEqual } = column;
+
+ let str: CifField['str']
+ let int: CifField['int']
+ let float: CifField['float']
+
+ switch (column.schema.valueType) {
+ case 'float':
+ case 'int':
+ str = row => { return '' + column.value(row); };
+ int = row => column.value(row);
+ float = row => column.value(row);
+ break
+ case 'str':
+ str = row => column.value(row);
+ int = row => { const v = column.value(row); return fastParseInt(v, 0, v.length) || 0; };
+ float = row => { const v = column.value(row); return fastParseFloat(v, 0, v.length) || 0; };
+ break
+ default:
+ throw new Error('unsupported')
+ }
+
+
+ return {
+ __array: void 0,
+ binaryEncoding: void 0,
+ isDefined: true,
+ rowCount,
+ str,
+ int,
+ float,
+ valueKind,
+ areValuesEqual,
+ toStringArray: params => ColumnHelpers.createAndFillArray(rowCount, str, params),
+ toIntArray: params => ColumnHelpers.createAndFillArray(rowCount, int, params),
+ toFloatArray: params => ColumnHelpers.createAndFillArray(rowCount, float, params)
+ }
+ }
}
export function getTensor(category: CifCategory, field: string, space: Tensor.Space, row: number, zeroIndexed: boolean): Tensor.Data {
diff --git a/src/mol-io/reader/gro/parser.ts b/src/mol-io/reader/gro/parser.ts
index 0367a3ee8..1a181f1a6 100644
--- a/src/mol-io/reader/gro/parser.ts
+++ b/src/mol-io/reader/gro/parser.ts
@@ -155,10 +155,8 @@ async function parseInternal(data: string, ctx: RuntimeContext): Promise<Result<
return Result.success(result);
}
-export function parse(data: string) {
+export function parseGRO(data: string) {
return Task.create<Result<Schema.GroFile>>('Parse GRO', async ctx => {
return await parseInternal(data, ctx);
});
}
-
-export default parse;
\ No newline at end of file
diff --git a/src/mol-model-formats/structure/gro.ts b/src/mol-model-formats/structure/gro.ts
index 1e04fb1e7..bd75839a3 100644
--- a/src/mol-model-formats/structure/gro.ts
+++ b/src/mol-model-formats/structure/gro.ts
@@ -1,154 +1,82 @@
-// TODO: make this work when the time comes.
-// /**
-// * Copyright (c) 2018 mol* contributors, licensed under MIT, See LICENSE file for more info.
-// *
-// * @author Alexander Rose <alexander.rose@weirdbyte.de>
-// */
-
-// import { Column, Table } from 'mol-data/db';
-// import { Interval, Segmentation } from 'mol-data/int';
-// import { mmCIF_Schema as mmCIF } from 'mol-io/reader/cif/schema/mmcif';
-// import { Atoms } from 'mol-io/reader/gro/schema';
-// import UUID from 'mol-util/uuid';
-// import Format from '../format';
-// import Model from '../model';
-// import { AtomicConformation, AtomicData, AtomicSegments, AtomsSchema, ChainsSchema, ResiduesSchema } from '../properties/atomic';
-// import { CoarseHierarchy } from '../properties/coarse';
-// import { Entities } from '../properties/common';
-// import Sequence from '../properties/sequence';
-// import { ModelSymmetry } from '../properties/symmetry';
-// import { guessElement } from '../properties/utils/guess-element';
-// import { getAtomicKeys } from '../properties/utils/keys';
-// import { ElementSymbol } from '../types';
-
-// import gro_Format = Format.gro
-
-// type HierarchyOffsets = { residues: ArrayLike<number>, chains: ArrayLike<number> }
-
-// function findHierarchyOffsets(atomsData: Atoms, bounds: Interval) {
-// const start = Interval.start(bounds), end = Interval.end(bounds);
-// const residues = [start], chains = [start];
-
-// const { residueName, residueNumber } = atomsData;
-
-// for (let i = start + 1; i < end; i++) {
-// const newResidue = !residueNumber.areValuesEqual(i - 1, i)
-// || !residueName.areValuesEqual(i - 1, i);
-// console.log(residueName.value(i - 1), residueName.value(i), residueNumber.value(i - 1), residueNumber.value(i), newResidue)
-// if (newResidue) residues[residues.length] = i;
-// }
-// console.log(residues, residues.length)
-// return { residues, chains };
-// }
-
-// function guessElementSymbol (value: string) {
-// return ElementSymbol(guessElement(value));
-// }
-
-// function createHierarchyData(atomsData: Atoms, offsets: HierarchyOffsets): AtomicData {
-// console.log(atomsData.atomName)
-// const atoms = Table.ofColumns(AtomsSchema, {
-// type_symbol: Column.ofArray({ array: Column.mapToArray(atomsData.atomName, guessElementSymbol), schema: Column.Schema.Aliased<ElementSymbol>(Column.Schema.str) }),
-// label_atom_id: atomsData.atomName,
-// auth_atom_id: atomsData.atomName,
-// label_alt_id: Column.Undefined(atomsData.count, Column.Schema.str),
-// pdbx_formal_charge: Column.Undefined(atomsData.count, Column.Schema.int)
-// });
-
-// const residues = Table.view(Table.ofColumns(ResiduesSchema, {
-// group_PDB: Column.Undefined(atomsData.count, Column.Schema.Aliased<'ATOM' | 'HETATM'>(Column.Schema.str)),
-// label_comp_id: atomsData.residueName,
-// auth_comp_id: atomsData.residueName,
-// label_seq_id: atomsData.residueNumber,
-// auth_seq_id: atomsData.residueNumber,
-// pdbx_PDB_ins_code: Column.Undefined(atomsData.count, Column.Schema.str),
-// }), ResiduesSchema, offsets.residues);
-// // Optimize the numeric columns
-// Table.columnToArray(residues, 'label_seq_id', Int32Array);
-// Table.columnToArray(residues, 'auth_seq_id', Int32Array);
-
-// // const chains = Table.ofColumns(Hierarchy.ChainsSchema, {
-// // label_asym_id: Column.ofConst('A', atomsData.count, Column.Schema.str),
-// // auth_asym_id: Column.ofConst('A', atomsData.count, Column.Schema.str),
-// // label_entity_id: Column.Undefined(atomsData.count, Column.Schema.str)
-// // });
-
-// const chains = Table.ofUndefinedColumns(ChainsSchema, 0);
-
-// return { atoms, residues, chains };
-// }
-
-// function getConformation(atoms: Atoms): AtomicConformation {
-// return {
-// id: UUID.create(),
-// atomId: atoms.atomNumber,
-// occupancy: Column.Undefined(atoms.count, Column.Schema.int),
-// B_iso_or_equiv: Column.Undefined(atoms.count, Column.Schema.float),
-// x: Column.mapToArray(atoms.x, x => x * 10, Float32Array),
-// y: Column.mapToArray(atoms.y, y => y * 10, Float32Array),
-// z: Column.mapToArray(atoms.z, z => z * 10, Float32Array)
-// }
-// }
-
-// function isHierarchyDataEqual(a: AtomicData, b: AtomicData) {
-// // need to cast because of how TS handles type resolution for interfaces https://github.com/Microsoft/TypeScript/issues/15300
-// return Table.areEqual(a.residues as Table<ResiduesSchema>, b.residues as Table<ResiduesSchema>)
-// && Table.areEqual(a.atoms as Table<AtomsSchema>, b.atoms as Table<AtomsSchema>)
-// }
-
-// function createModel(format: gro_Format, modelNum: number, previous?: Model): Model {
-// const structure = format.data.structures[modelNum];
-// const bounds = Interval.ofBounds(0, structure.atoms.count);
-
-// const hierarchyOffsets = findHierarchyOffsets(structure.atoms, bounds);
-// const hierarchyData = createHierarchyData(structure.atoms, hierarchyOffsets);
-
-// if (previous && isHierarchyDataEqual(previous.atomicHierarchy, hierarchyData)) {
-// return {
-// ...previous,
-// atomicConformation: getConformation(structure.atoms)
-// };
-// }
-
-// const hierarchySegments: AtomicSegments = {
-// residueSegments: Segmentation.ofOffsets(hierarchyOffsets.residues, bounds),
-// chainSegments: Segmentation.ofOffsets(hierarchyOffsets.chains, bounds),
-// }
-
-// // TODO: create a better mock entity
-// const entityTable = Table.ofRows<mmCIF['entity']>(mmCIF.entity, [{
-// id: '0',
-// src_method: 'syn',
-// type: 'polymer',
-// pdbx_number_of_molecules: 1
-// }]);
-
-// const entities: Entities = { data: entityTable, getEntityIndex: Column.createIndexer(entityTable.id) };
-
-// const hierarchyKeys = getAtomicKeys(hierarchyData, entities, hierarchySegments);
-// const atomicHierarchy = { ...hierarchyData, ...hierarchyKeys, ...hierarchySegments };
-// return {
-// id: UUID.create(),
-// sourceData: format,
-// modelNum,
-// atomicHierarchy,
-// entities,
-// sequence: Sequence.fromAtomicHierarchy(atomicHierarchy),
-// atomicConformation: getConformation(structure.atoms),
-// coarseHierarchy: CoarseHierarchy.Empty,
-// coarseConformation: void 0 as any,
-// symmetry: ModelSymmetry.Default
-// };
-// }
-
-// function buildModels(format: gro_Format): ReadonlyArray<Model> {
-// const models: Model[] = [];
-
-// format.data.structures.forEach((_, i) => {
-// const model = createModel(format, i, models.length > 0 ? models[models.length - 1] : void 0);
-// models.push(model);
-// });
-// return models;
-// }
-
-// export default buildModels;
+/**
+ * Copyright (c) 2019 mol* contributors, licensed under MIT, See LICENSE file for more info.
+ *
+ * @author Alexander Rose <alexander.rose@weirdbyte.de>
+ */
+
+import { Model } from 'mol-model/structure/model';
+import { Task } from 'mol-task';
+import { ModelFormat } from './format';
+import { _parse_mmCif } from './mmcif/parser';
+import { GroFile, GroAtoms } from 'mol-io/reader/gro/schema';
+import { CifCategory, CifField } from 'mol-io/reader/cif';
+import { Column } from 'mol-data/db';
+import { mmCIF_Schema } from 'mol-io/reader/cif/schema/mmcif';
+
+// TODO multi model files
+// TODO seperate chains
+// TODO better entity handling
+// TODO improve performance
+
+function _entity(): { [K in keyof mmCIF_Schema['entity']]?: CifField } {
+ return {
+ id: CifField.ofStrings(['1', '2', '3']),
+ type: CifField.ofStrings(['polymer', 'non-polymer', 'water'])
+ }
+}
+
+function _atom_site(atoms: GroAtoms): { [K in keyof mmCIF_Schema['atom_site']]?: CifField } {
+ const auth_asym_id = CifField.ofColumn(Column.Undefined(atoms.count, Column.Schema.str))
+ const auth_atom_id = CifField.ofColumn(atoms.atomName)
+ const auth_comp_id = CifField.ofColumn(atoms.residueName)
+ const auth_seq_id = CifField.ofColumn(atoms.residueNumber)
+
+ return {
+ auth_asym_id,
+ auth_atom_id,
+ auth_comp_id,
+ auth_seq_id,
+ B_iso_or_equiv: CifField.ofColumn(Column.Undefined(atoms.count, Column.Schema.float)),
+ Cartn_x: CifField.ofNumbers(Column.mapToArray(atoms.x, x => x * 10, Float32Array)),
+ Cartn_y: CifField.ofNumbers(Column.mapToArray(atoms.y, y => y * 10, Float32Array)),
+ Cartn_z: CifField.ofNumbers(Column.mapToArray(atoms.z, z => z * 10, Float32Array)),
+ group_PDB: CifField.ofColumn(Column.Undefined(atoms.count, Column.Schema.str)),
+ id: CifField.ofColumn(atoms.atomNumber),
+
+ label_alt_id: CifField.ofColumn(Column.Undefined(atoms.count, Column.Schema.str)),
+
+ label_asym_id: auth_asym_id,
+ label_atom_id: auth_atom_id,
+ label_comp_id: auth_comp_id,
+ label_seq_id: auth_seq_id,
+ label_entity_id: CifField.ofColumn(Column.ofConst('1', atoms.count, Column.Schema.str)),
+
+ occupancy: CifField.ofColumn(Column.ofConst(1, atoms.count, Column.Schema.float)),
+ type_symbol: CifField.ofColumn(Column.Undefined(atoms.count, Column.Schema.str)),
+
+ pdbx_PDB_ins_code: CifField.ofColumn(Column.Undefined(atoms.count, Column.Schema.str)),
+ pdbx_PDB_model_num: CifField.ofColumn(Column.ofConst('1', atoms.count, Column.Schema.str)),
+ }
+}
+
+async function groToMmCif(gro: GroFile) {
+ const categories = {
+ entity: CifCategory.ofFields('entity', _entity()),
+ atom_site: CifCategory.ofFields('atom_site', _atom_site(gro.structures[0].atoms))
+ } as any;
+
+ return {
+ header: 'GRO',
+ categoryNames: Object.keys(categories),
+ categories
+ };
+}
+
+export function trajectoryFromGRO(gro: GroFile): Task<Model.Trajectory> {
+ return Task.create('Parse GRO', async ctx => {
+ await ctx.update('Converting to mmCIF');
+ const cif = await groToMmCif(gro);
+ const format = ModelFormat.mmCIF(cif);
+ return _parse_mmCif(format, ctx);
+ })
+}
diff --git a/src/mol-model-formats/structure/pdb.ts b/src/mol-model-formats/structure/pdb.ts
index a86958769..cce4dc869 100644
--- a/src/mol-model-formats/structure/pdb.ts
+++ b/src/mol-model-formats/structure/pdb.ts
@@ -1,5 +1,5 @@
/**
- * Copyright (c) 2017-2018 mol* contributors, licensed under MIT, See LICENSE file for more info.
+ * Copyright (c) 2019 mol* contributors, licensed under MIT, See LICENSE file for more info.
*
* @author David Sehnal <david.sehnal@gmail.com>
*/
diff --git a/src/mol-plugin/state/actions/data-format.ts b/src/mol-plugin/state/actions/data-format.ts
index eb9ad4612..7e4bbd656 100644
--- a/src/mol-plugin/state/actions/data-format.ts
+++ b/src/mol-plugin/state/actions/data-format.ts
@@ -12,7 +12,7 @@ import { PluginStateObject } from '../objects';
import { ParamDefinition as PD } from 'mol-util/param-definition';
import { Ccp4Provider, Dsn6Provider, DscifProvider } from './volume';
import { StateTransforms } from '../transforms';
-import { MmcifProvider, PdbProvider } from './structure';
+import { MmcifProvider, PdbProvider, GroProvider } from './structure';
export class DataFormatRegistry<D extends PluginStateObject.Data.Binary | PluginStateObject.Data.String> {
private _list: { name: string, provider: DataFormatProvider<D> }[] = []
@@ -56,6 +56,7 @@ export class DataFormatRegistry<D extends PluginStateObject.Data.Binary | Plugin
this.add('ccp4', Ccp4Provider)
this.add('dscif', DscifProvider)
this.add('dsn6', Dsn6Provider)
+ this.add('gro', GroProvider)
this.add('mmcif', MmcifProvider)
this.add('pdb', PdbProvider)
};
diff --git a/src/mol-plugin/state/actions/structure.ts b/src/mol-plugin/state/actions/structure.ts
index 53696f1eb..0a5e93d99 100644
--- a/src/mol-plugin/state/actions/structure.ts
+++ b/src/mol-plugin/state/actions/structure.ts
@@ -22,7 +22,7 @@ export const MmcifProvider: DataFormatProvider<any> = {
description: 'mmCIF',
stringExtensions: ['cif', 'mmcif', 'mcif'],
binaryExtensions: ['bcif'],
- isApplicable: (info: FileInfo, data: Uint8Array) => {
+ isApplicable: (info: FileInfo, data: Uint8Array | string) => {
return info.ext === 'cif' || info.ext === 'mmcif' || info.ext === 'mcif' || info.ext === 'bcif'
},
getDefaultBuilder: (ctx: PluginContext, data: StateBuilder.To<PluginStateObject.Data.Binary | PluginStateObject.Data.String>, state: State) => {
@@ -38,17 +38,33 @@ export const PdbProvider: DataFormatProvider<any> = {
description: 'PDB',
stringExtensions: ['pdb', 'ent'],
binaryExtensions: [],
- isApplicable: (info: FileInfo, data: Uint8Array) => {
+ isApplicable: (info: FileInfo, data: string) => {
return info.ext === 'pdb' || info.ext === 'ent'
},
getDefaultBuilder: (ctx: PluginContext, data: StateBuilder.To<PluginStateObject.Data.String>, state: State) => {
- return Task.create('mmCIF default builder', async taskCtx => {
+ return Task.create('PDB default builder', async taskCtx => {
const traj = createModelTree(data, 'pdb');
await state.updateTree(createStructureTree(ctx, traj, false)).runInContext(taskCtx)
})
}
}
+export const GroProvider: DataFormatProvider<any> = {
+ label: 'GRO',
+ description: 'GRO',
+ stringExtensions: ['gro'],
+ binaryExtensions: [],
+ isApplicable: (info: FileInfo, data: string) => {
+ return info.ext === 'gro'
+ },
+ getDefaultBuilder: (ctx: PluginContext, data: StateBuilder.To<PluginStateObject.Data.String>, state: State) => {
+ return Task.create('GRO default builder', async taskCtx => {
+ const traj = createModelTree(data, 'gro');
+ await state.updateTree(createStructureTree(ctx, traj, false)).runInContext(taskCtx)
+ })
+ }
+}
+
//
export { DownloadStructure };
@@ -111,10 +127,22 @@ const DownloadStructure = StateAction.build({
return state.updateTree(createStructureTree(ctx, traj, params.source.params.supportProps));
});
-function createModelTree(b: StateBuilder.To<PluginStateObject.Data.Binary | PluginStateObject.Data.String>, format: 'pdb' | 'cif' = 'cif') {
- const parsed = format === 'cif'
- ? b.apply(StateTransforms.Data.ParseCif, void 0, { props: { isGhost: true }}).apply(StateTransforms.Model.TrajectoryFromMmCif, void 0, { props: { isGhost: true }})
- : b.apply(StateTransforms.Model.TrajectoryFromPDB, void 0, { props: { isGhost: true }});
+function createModelTree(b: StateBuilder.To<PluginStateObject.Data.Binary | PluginStateObject.Data.String>, format: 'pdb' | 'cif' | 'gro' = 'cif') {
+ let parsed: StateBuilder.To<PluginStateObject.Molecule.Trajectory>
+ switch (format) {
+ case 'cif':
+ parsed = b.apply(StateTransforms.Data.ParseCif, void 0, { props: { isGhost: true }})
+ .apply(StateTransforms.Model.TrajectoryFromMmCif, void 0, { props: { isGhost: true }})
+ break
+ case 'pdb':
+ parsed = b.apply(StateTransforms.Model.TrajectoryFromPDB, void 0, { props: { isGhost: true }});
+ break
+ case 'gro':
+ parsed = b.apply(StateTransforms.Model.TrajectoryFromGRO, void 0, { props: { isGhost: true }});
+ break
+ default:
+ throw new Error('unsupported format')
+ }
return parsed.apply(StateTransforms.Model.ModelFromTrajectory, { modelIndex: 0 });
}
diff --git a/src/mol-plugin/state/transforms/model.ts b/src/mol-plugin/state/transforms/model.ts
index 7131485bb..67b29a021 100644
--- a/src/mol-plugin/state/transforms/model.ts
+++ b/src/mol-plugin/state/transforms/model.ts
@@ -20,9 +20,12 @@ import { RuntimeContext, Task } from 'mol-task';
import { ParamDefinition as PD } from 'mol-util/param-definition';
import { stringToWords } from 'mol-util/string';
import { PluginStateObject as SO, PluginStateTransform } from '../objects';
+import { trajectoryFromGRO } from 'mol-model-formats/structure/gro';
+import { parseGRO } from 'mol-io/reader/gro/parser';
export { TrajectoryFromMmCif };
export { TrajectoryFromPDB };
+export { TrajectoryFromGRO };
export { ModelFromTrajectory };
export { StructureFromModel };
export { StructureAssemblyFromModel };
@@ -62,7 +65,6 @@ const TrajectoryFromMmCif = PluginStateTransform.BuiltIn({
}
});
-
type TrajectoryFromPDB = typeof TrajectoryFromPDB
const TrajectoryFromPDB = PluginStateTransform.BuiltIn({
name: 'trajectory-from-pdb',
@@ -81,6 +83,23 @@ const TrajectoryFromPDB = PluginStateTransform.BuiltIn({
}
});
+type TrajectoryFromGRO = typeof TrajectoryFromGRO
+const TrajectoryFromGRO = PluginStateTransform.BuiltIn({
+ name: 'trajectory-from-gro',
+ display: { name: 'Parse GRO', description: 'Parse GRO string and create trajectory.' },
+ from: [SO.Data.String],
+ to: SO.Molecule.Trajectory
+})({
+ apply({ a }) {
+ return Task.create('Parse GRO', async ctx => {
+ const parsed = await parseGRO(a.data).runInContext(ctx);
+ if (parsed.isError) throw new Error(parsed.message);
+ const models = await trajectoryFromGRO(parsed.result).runInContext(ctx);
+ const props = { label: models[0].label, description: `${models.length} model${models.length === 1 ? '' : 's'}` };
+ return new SO.Molecule.Trajectory(models, props);
+ });
+ }
+});
const plus1 = (v: number) => v + 1, minus1 = (v: number) => v - 1;
type ModelFromTrajectory = typeof ModelFromTrajectory
--
GitLab