From a81bcf4311624b15525cfe4b376c667fbf40cbdd Mon Sep 17 00:00:00 2001
From: David Sehnal <david.sehnal@gmail.com>
Date: Sat, 2 Mar 2019 12:57:58 +0100
Subject: [PATCH] mol-model: added sourceIndex to AtomicHierarchy
---
package-lock.json | Bin 609362 -> 607562 bytes
.../structure/mmcif/atomic.ts | 12 +++--
src/mol-model-formats/structure/mmcif/ihm.ts | 1 +
.../structure/mmcif/parser.ts | 41 +++++++++++++-----
src/mol-model-formats/structure/mmcif/sort.ts | 9 +++-
.../model/properties/atomic/hierarchy.ts | 9 +++-
6 files changed, 50 insertions(+), 22 deletions(-)
diff --git a/package-lock.json b/package-lock.json
index 4444de29a715c5abccb6351afde6e97dd2a96c96..00f2081bcfdf7d63fc7301f534af8309da642912 100644
GIT binary patch
delta 1504
zcmcaKL-kab>IQYM$qRJlHoJH|+dKUME2GSIR}Mz$rpXH~h)r)`XJef%@QYP+`jsWD
zg3~{6F^O*9-o<F{1(My)dX>?bV|qaVi^O*SyNm)Z(-lmaB(|3+GOdo7zCnXkV)`ys
zX3^;a^O*!DCuDO^SD4PhGToqtDSEn&F`MxAA5~28jFT_?7oT2W#l$*&LJd>?WQPsX
z)8FNBXl#$KWzvC2?yqA?0ISmK=a8D7Th9~>=KZzd;GeEl#KF5=pn+)yH%M&zoo*%$
zMz9%q*32r~9eSC>Al$w_ra2%L(-RgjvP}Qp%po$}po@`j`hQ1e!Rh)9Os3Nn%s92C
zPuR`MI$eJalhX9pD_Oaw3)nNWOcwYf0nue+&nz<i!yZPy$q%ORPEV*~@|b>LI+Nyf
z0Xt^y>Hk5J4Yf=X(_`xxIj1`~F!N47Fo}_Gdb|U(;PeX}Os3lpOk{H706Ri=9y|NA
znQWZX=a{i^PT%0g%rg1GKfdV`Y*~b+KbXlBJzZ-ehcMWx>HZTr1VKS&3ig%Wc2<?i
z2mbL)xAI|@nl8}5R5iWpH9OCAcOPb6Fn9ZjIZTo4(+~KtOH6-d#>_X}U=9Z_$g2v|
z4H}rDw!dD&w1a8+f#pn0lN~mMPUrv6%sT!3A`aQ<0#{j~O3tlfs$-sf;lJJV3Gdmt
zrz?oDv4ZrfO@1J&yj^xZ(|z{o8#b})Pe0qux^?=2Tg(#Mzi(%<&;X@}>H8Njc2D=X
z&&0d^-UFuDF4Gmdnfa$b$l?&(J~x}gmkGkZpwBEYy|0f+X!_h5Cf?}*_RK=lqw_d|
zr$1=oP}u%HkE5LtNnt`Q6Yul~TFmiCV)fe0l}OxmtJwLsYZY;n^Fjm{gs=#1|8L2x
z&p5q7msx&$o;7nGBgpj2VEJiw%styDv~aL6O>Zz@R-V2gkePk^{WgwhPO$WJu=~zi
zva@Y(n82Y3F=P2ZZ)PqeX@#}yEYmO4aBy!on8R_L8Ent=+sm2Mr%h(zn4Z(aB)+|N
z0mpNQ=ILiO7!{{0_%NwV_cLZ@+x~tL$2Jxu!{XwZxu(Bg&H?tKar^!x=Jx$bEZg@d
zu?ELa7qDlM*sghqRYV3B!Xn!<?y<IWfVkV)p0iFAo9@udsJVUrUsgZS=>n6ORX6`w
zo5VPMK?tMz<c9Uu(-k@y*|z@{VVfWTj=1e-yID6dPB-|-$g|z=AoDS%?SF5u#xZW6
zaE#fQar+q^wxgQU6$;oSwyy|dlYa}=0*dQ{(@(i`@=xzq=j5EeeiD<wWPu~r({HPD
z%4|1vWtSD0J|TxuV*0#~jFQ`%LfM^iK&}HNhq`I4?2^-;2Qf-c_kYU5vwiv@_LFSW
z1^%&1Y%f2_{&XQE7)Fx_Nl9e-N`DT%?Hg8b{Hg?p^L=Gb_U$K>IAz7a+=RufV%vo*
HIUR%n)*d77
delta 1330
zcmX@rrFv<G>IQYM=@Tw9iA}CIV&ClQ^=vPgH+{kv4$ke~9E{>klLdIir~P7O-Co<p
zSmZVR!gfZ9?QB;WjX1yxxBK5^6m)?w%N3c{L`)ZGWHQ-qRmH@_1ZGYT&E$}up0J!<
zaC&DZ2mkc{txUPoe`IoKOkdE##Il{QmPvyVqR_XFDFMRTtH&la{k#S<@AMC~Ol}ad
zD1T=D=`otjywk3*C`>P?Vv5>+ubxSc8zP%t#UVWXf)<Cs^u8(%{^|YgOfl1UTd})M
zYh%ipE?~lGGrj66lk)TiR#p+P-sula*;%Lan=-Rc=kH}ooPJ?0lg0E0s?0*uA6PPr
zOg>j*Gu^?8S$J9>lRCtL2d3;S)6cGCVxPXoikWx%gBoUm=>c}kQ2!XLVlvzQyN@ZG
z5n|E?Csx+U=egLXA86nZ-adZ<6EDZK*-WO>Wp=XhPxtEL;GO<p2Gb*m+Pp3%gYCSt
zm>M`BHfaSh%WOZeh{=c<EWe$18Pij!u?|6utkX9rvYAf5pw7rT{lWwef$cFXm{?gL
zIxE&N#e!HG+aIi9`p!OmLIS(^bOQ~>C)+P?WwOwiu28`&Io-jGy=MA>nT$N!SKnh=
z=`uYaox^c@oIW$h_8MhoSElI>q0A1TkZ75{z>e8r`;SbH4n~OJ{A`XK2y4%DCUE@3
zL%0SDnE1EHXfl`cLil>~nAo=GR&i)Cg3Xw|rjtWpx`GL($n*tX%qrXeRdc98WzJbL
zZ-MC9;Ka%@JwJ*W9KMARefOf6t+v0e<5&$*z1_i<IgJydobN6R+jfU84n>I6@`i2>
zE|84l_61Wpt}=u5Z<h~a)@7P*;Kb}Wy?-l{0>t(W5zKp7KuR5^E#$}qF$|_ZxWcTm
z-FG#!7*o4#0&}}<0?T&W1XkNvNV1-|msL~-9(Y3AFWh8p=K%2~w&y)!ogfAYXZ|2|
z!R_ilS^Y%8x~2>KWais`pouk!ar+;BwjTlzRR$W2XQnGGV&a|NSIH(Y-NBSaa=JnX
z>rqII+*e|g(T8Z)+s!JoT`!1@<t>uN1U*Kc>8`?@{L{nkvC2+g5XNM_z14x;5^DGa
z5e~`i?*rJKazL&FNA=Y0_x7^CV*{JH{qYg@XA4IooEaM7wh7D+n6?+(<6vdoo*&J`
m&$xZUDh`pt=?`Q%<+dM`=9Co!F(sy-oX0A<oy&;RK^OpVug+Ki
diff --git a/src/mol-model-formats/structure/mmcif/atomic.ts b/src/mol-model-formats/structure/mmcif/atomic.ts
index b43be7fe7..38e39f5b6 100644
--- a/src/mol-model-formats/structure/mmcif/atomic.ts
+++ b/src/mol-model-formats/structure/mmcif/atomic.ts
@@ -16,11 +16,8 @@ import { ElementSymbol } from 'mol-model/structure/model/types';
import { Entities } from 'mol-model/structure/model/properties/common';
import { getAtomicRanges } from 'mol-model/structure/model/properties/utils/atomic-ranges';
import { getAtomicDerivedData } from 'mol-model/structure/model/properties/utils/atomic-derived';
-import { ModelFormat } from '../format';
-import mmCIF_Format = ModelFormat.mmCIF
import { FormatData } from './parser';
-
type AtomSite = mmCIF_Database['atom_site']
function findHierarchyOffsets(atom_site: AtomSite) {
@@ -45,13 +42,14 @@ function findHierarchyOffsets(atom_site: AtomSite) {
return { residues, chains };
}
-function createHierarchyData(atom_site: AtomSite, offsets: { residues: ArrayLike<number>, chains: ArrayLike<number> }): AtomicData {
+function createHierarchyData(atom_site: AtomSite, sourceIndex: Column<number>, offsets: { residues: ArrayLike<number>, chains: ArrayLike<number> }): AtomicData {
const atoms = Table.ofColumns(AtomsSchema, {
type_symbol: Column.ofArray({ array: Column.mapToArray(atom_site.type_symbol, ElementSymbol), schema: Column.Schema.Aliased<ElementSymbol>(Column.Schema.str) }),
label_atom_id: atom_site.label_atom_id,
auth_atom_id: atom_site.auth_atom_id,
label_alt_id: atom_site.label_alt_id,
- pdbx_formal_charge: atom_site.pdbx_formal_charge
+ pdbx_formal_charge: atom_site.pdbx_formal_charge,
+ sourceIndex
});
const residues = Table.view(atom_site, ResiduesSchema, offsets.residues);
// Optimize the numeric columns
@@ -80,9 +78,9 @@ function isHierarchyDataEqual(a: AtomicData, b: AtomicData) {
&& Table.areEqual(a.atoms as Table<AtomsSchema>, b.atoms as Table<AtomsSchema>)
}
-export function getAtomicHierarchyAndConformation(format: mmCIF_Format, atom_site: AtomSite, entities: Entities, formatData: FormatData, previous?: Model) {
+export function getAtomicHierarchyAndConformation(atom_site: AtomSite, sourceIndex: Column<number>, entities: Entities, formatData: FormatData, previous?: Model) {
const hierarchyOffsets = findHierarchyOffsets(atom_site);
- const hierarchyData = createHierarchyData(atom_site, hierarchyOffsets);
+ const hierarchyData = createHierarchyData(atom_site, sourceIndex, hierarchyOffsets);
if (previous && isHierarchyDataEqual(previous.atomicHierarchy, hierarchyData)) {
return {
diff --git a/src/mol-model-formats/structure/mmcif/ihm.ts b/src/mol-model-formats/structure/mmcif/ihm.ts
index 731af9e3a..405fec8cf 100644
--- a/src/mol-model-formats/structure/mmcif/ihm.ts
+++ b/src/mol-model-formats/structure/mmcif/ihm.ts
@@ -21,6 +21,7 @@ export interface IHMData {
model_name: string,
entities: Entities,
atom_site: mmCIF['atom_site'],
+ atom_site_sourceIndex: Column<number>,
ihm_sphere_obj_site: mmCIF['ihm_sphere_obj_site'],
ihm_gaussian_obj_site: mmCIF['ihm_gaussian_obj_site']
}
diff --git a/src/mol-model-formats/structure/mmcif/parser.ts b/src/mol-model-formats/structure/mmcif/parser.ts
index 17d4f3a5b..44835f9e9 100644
--- a/src/mol-model-formats/structure/mmcif/parser.ts
+++ b/src/mol-model-formats/structure/mmcif/parser.ts
@@ -169,8 +169,8 @@ function getFormatData(format: mmCIF_Format): FormatData {
}
}
-function createStandardModel(format: mmCIF_Format, atom_site: AtomSite, entities: Entities, formatData: FormatData, previous?: Model): Model {
- const atomic = getAtomicHierarchyAndConformation(format, atom_site, entities, formatData, previous);
+function createStandardModel(format: mmCIF_Format, atom_site: AtomSite, sourceIndex: Column<number>, entities: Entities, formatData: FormatData, previous?: Model): Model {
+ const atomic = getAtomicHierarchyAndConformation(atom_site, sourceIndex, entities, formatData, previous);
if (previous && atomic.sameAsPrevious) {
return {
...previous,
@@ -209,7 +209,7 @@ function createStandardModel(format: mmCIF_Format, atom_site: AtomSite, entities
}
function createModelIHM(format: mmCIF_Format, data: IHMData, formatData: FormatData): Model {
- const atomic = getAtomicHierarchyAndConformation(format, data.atom_site, data.entities, formatData);
+ const atomic = getAtomicHierarchyAndConformation(data.atom_site, data.atom_site_sourceIndex, data.entities, formatData);
const coarse = getIHMCoarse(data, formatData);
return {
@@ -255,8 +255,8 @@ async function readStandard(ctx: RuntimeContext, format: mmCIF_Format, formatDat
let modelStart = 0;
while (modelStart < atomCount) {
const modelEnd = findModelEnd(format.data.atom_site.pdbx_PDB_model_num, modelStart);
- const atom_site = await sortAtomSite(ctx, format.data.atom_site, modelStart, modelEnd);
- const model = createStandardModel(format, atom_site, entities, formatData, models.length > 0 ? models[models.length - 1] : void 0);
+ const { atom_site, sourceIndex } = await sortAtomSite(ctx, format.data.atom_site, modelStart, modelEnd);
+ const model = createStandardModel(format, atom_site, sourceIndex, entities, formatData, models.length > 0 ? models[models.length - 1] : void 0);
attachProps(model);
models.push(model);
modelStart = modelEnd;
@@ -265,14 +265,17 @@ async function readStandard(ctx: RuntimeContext, format: mmCIF_Format, formatDat
}
function splitTable<T extends Table<any>>(table: T, col: Column<number>) {
- const ret = new Map<number, T>()
+ const ret = new Map<number, { table: T, start: number, end: number }>()
const rowCount = table._rowCount;
let modelStart = 0;
while (modelStart < rowCount) {
const modelEnd = findModelEnd(col, modelStart);
const id = col.value(modelStart);
- const window = Table.window(table, table._schema, modelStart, modelEnd) as T;
- ret.set(id, window);
+ ret.set(id, {
+ table: Table.window(table, table._schema, modelStart, modelEnd) as T,
+ start: modelStart,
+ end: modelEnd
+ });
modelStart = modelEnd;
}
return ret;
@@ -286,8 +289,9 @@ async function readIHM(ctx: RuntimeContext, format: mmCIF_Format, formatData: Fo
throw new Error('expected _atom_site.ihm_model_id to be defined')
}
- // TODO: will IHM require sorting or will we trust it?
const atom_sites = splitTable(format.data.atom_site, format.data.atom_site.ihm_model_id);
+ // TODO: will coarse IHM records require sorting or will we trust it?
+ // ==> Probably implement a sort as as well and store the sourceIndex same as with atomSite
const sphere_sites = splitTable(format.data.ihm_sphere_obj_site, format.data.ihm_sphere_obj_site.model_id);
const gauss_sites = splitTable(format.data.ihm_gaussian_obj_site, format.data.ihm_gaussian_obj_site.model_id);
@@ -296,13 +300,26 @@ async function readIHM(ctx: RuntimeContext, format: mmCIF_Format, formatData: Fo
const { model_id, model_name } = ihm_model_list;
for (let i = 0; i < ihm_model_list._rowCount; i++) {
const id = model_id.value(i);
+
+ let atom_site, atom_site_sourceIndex;
+ if (atom_sites.has(id)) {
+ const e = atom_sites.get(id)!;
+ const { atom_site: sorted, sourceIndex } = await sortAtomSite(ctx, e.table, e.start, e.end);
+ atom_site = sorted;
+ atom_site_sourceIndex = sourceIndex;
+ } else {
+ atom_site = Table.window(format.data.atom_site, format.data.atom_site._schema, 0, 0);
+ atom_site_sourceIndex = Column.ofIntArray([]);
+ }
+
const data: IHMData = {
model_id: id,
model_name: model_name.value(i),
entities: entities,
- atom_site: atom_sites.has(id) ? atom_sites.get(id)! : Table.window(format.data.atom_site, format.data.atom_site._schema, 0, 0),
- ihm_sphere_obj_site: sphere_sites.has(id) ? sphere_sites.get(id)! : Table.window(format.data.ihm_sphere_obj_site, format.data.ihm_sphere_obj_site._schema, 0, 0),
- ihm_gaussian_obj_site: gauss_sites.has(id) ? gauss_sites.get(id)! : Table.window(format.data.ihm_gaussian_obj_site, format.data.ihm_gaussian_obj_site._schema, 0, 0)
+ atom_site,
+ atom_site_sourceIndex,
+ ihm_sphere_obj_site: sphere_sites.has(id) ? sphere_sites.get(id)!.table : Table.window(format.data.ihm_sphere_obj_site, format.data.ihm_sphere_obj_site._schema, 0, 0),
+ ihm_gaussian_obj_site: gauss_sites.has(id) ? gauss_sites.get(id)!.table : Table.window(format.data.ihm_gaussian_obj_site, format.data.ihm_gaussian_obj_site._schema, 0, 0)
};
const model = createModelIHM(format, data, formatData);
attachProps(model);
diff --git a/src/mol-model-formats/structure/mmcif/sort.ts b/src/mol-model-formats/structure/mmcif/sort.ts
index 7cbee9e5b..868cd5dd8 100644
--- a/src/mol-model-formats/structure/mmcif/sort.ts
+++ b/src/mol-model-formats/structure/mmcif/sort.ts
@@ -9,6 +9,8 @@ import { createRangeArray, makeBuckets } from 'mol-data/util';
import { Column, Table } from 'mol-data/db';
import { RuntimeContext } from 'mol-task';
+export type SortedAtomSite = mmCIF_Database['atom_site'] & { sourceIndex: Column<number> }
+
function isIdentity(xs: ArrayLike<number>) {
for (let i = 0, _i = xs.length; i < _i; i++) {
if (xs[i] !== i) return false;
@@ -36,8 +38,11 @@ export async function sortAtomSite(ctx: RuntimeContext, atom_site: mmCIF_Databas
}
if (isIdentity(indices) && indices.length === atom_site._rowCount) {
- return atom_site;
+ return { atom_site, sourceIndex: Column.ofIntArray(indices) };
}
- return Table.view(atom_site, atom_site._schema, indices) as mmCIF_Database['atom_site'];
+ return {
+ atom_site: Table.view(atom_site, atom_site._schema, indices) as mmCIF_Database['atom_site'],
+ sourceIndex: Column.ofIntArray(indices)
+ };
}
\ No newline at end of file
diff --git a/src/mol-model/structure/model/properties/atomic/hierarchy.ts b/src/mol-model/structure/model/properties/atomic/hierarchy.ts
index 33c599324..7262fe28a 100644
--- a/src/mol-model/structure/model/properties/atomic/hierarchy.ts
+++ b/src/mol-model/structure/model/properties/atomic/hierarchy.ts
@@ -38,7 +38,14 @@ export const AtomsSchema = {
* The net integer charge assigned to this atom.
* This is the formal charge assignment normally found in chemical diagrams.
*/
- pdbx_formal_charge: mmCIF.atom_site.pdbx_formal_charge
+ pdbx_formal_charge: mmCIF.atom_site.pdbx_formal_charge,
+
+ /**
+ * The index of this atom in the input data.
+ * Required because of sorting of atoms.
+ */
+ sourceIndex: Column.Schema.int
+
// id, occupancy and B_iso_or_equiv are part of conformation
};
--
GitLab