{
"cells": [
{
"cell_type": "markdown",
"metadata": {},
"source": [
"![\"Вернуться](\"https://pollytikhonova.github.io/bioinformatics_homeworks/GitHub-Mark-32px.png\" "\\"Вернуться")
\n",
"Tikhonova Polina. Homework 7.
"
]
},
{
"cell_type": "code",
"execution_count": 1,
"metadata": {
"collapsed": true
},
"outputs": [],
"source": [
"import numpy as np\n",
"import copy\n",
"\n",
"# Отображение структур\n",
"import IPython.display\n",
"import ipywidgets\n",
"from IPython.display import display,display_svg,SVG,Image, HTML\n",
"\n",
"# Open Drug Discovery Toolkit\n",
"import oddt\n",
"import oddt.docking\n",
"import oddt.interactions\n",
"\n",
"# Органика\n",
"from rdkit.Chem import Draw\n",
"from rdkit.Chem.Draw import IPythonConsole\n",
"\n",
"import pmx # Модуль для манипулирования pdb "
]
},
{
"cell_type": "code",
"execution_count": 2,
"metadata": {},
"outputs": [
{
"name": "stdout",
"output_type": "stream",
"text": [
"\n",
"\n",
"\n",
"\n",
"\n",
"\n",
"\n",
"\n",
"\n",
"\n",
"\n",
"\n",
"\n",
"\n",
"\n"
]
}
],
"source": [
"pdb=pmx.Model('with_ligand.B99990001.pdb')\n",
"\n",
"for r in pdb.residues[135:]:\n",
" print r #посмотрим остатки"
]
},
{
"cell_type": "code",
"execution_count": 3,
"metadata": {},
"outputs": [
{
"data": {
"text/html": [
"New pdb"
],
"text/plain": [
""
]
},
"metadata": {},
"output_type": "display_data"
},
{
"name": "stdout",
"output_type": "stream",
"text": [
"\n",
"\n",
"\n",
"\n",
"\n",
"\n",
"\n",
"\n",
"\n",
"\n",
"\n",
"\n"
]
},
{
"data": {
"text/html": [
"Ligand"
],
"text/plain": [
""
]
},
"metadata": {},
"output_type": "display_data"
},
{
"name": "stdout",
"output_type": "stream",
"text": [
"\n",
"\n",
"\n"
]
}
],
"source": [
"# создание объектов белок и лиганда\n",
"newpdb = pdb.copy()\n",
"for r in newpdb.residues[-3:]:\n",
" newpdb.remove_residue(r)\n",
"lig = pdb.copy()\n",
"del lig.residues[:-3]\n",
"for a in lig.atoms:\n",
" a.x\n",
"display(HTML('New pdb'))\n",
"for r in newpdb.residues[135:]:\n",
" print r #посмотрим остатки\n",
" \n",
"display(HTML('Ligand'))\n",
"for r in lig.residues:\n",
" print r #посмотрим остатки"
]
},
{
"cell_type": "code",
"execution_count": 4,
"metadata": {
"collapsed": true
},
"outputs": [],
"source": [
"coordinates = [a.x for a in lig.atoms]\n",
"center = np.mean(coordinates, axis=0)"
]
},
{
"cell_type": "code",
"execution_count": 5,
"metadata": {
"collapsed": true
},
"outputs": [],
"source": [
"newpdb.writePDB('my_prot.pdb')"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"### Подготовка белка для докинга"
]
},
{
"cell_type": "code",
"execution_count": 6,
"metadata": {},
"outputs": [
{
"name": "stdout",
"output_type": "stream",
"text": [
"Is it the first mol in 1lmp is protein? False :) and MW of this mol is: 16831.041\n"
]
}
],
"source": [
"prot = oddt.toolkit.readfile('pdb','my_prot.pdb').next()\n",
"\n",
"prot.OBMol.AddPolarHydrogens()\n",
"prot.OBMol.AutomaticPartialCharge()\n",
"\n",
"print 'Is it the first mol in 1lmp is protein?',prot.protein,':) and MW of this mol is:', prot.molwt "
]
},
{
"cell_type": "code",
"execution_count": 7,
"metadata": {},
"outputs": [
{
"data": {
"image/svg+xml": [
""
]
},
"metadata": {},
"output_type": "display_data"
},
{
"data": {
"image/svg+xml": [
""
]
},
"metadata": {},
"output_type": "display_data"
},
{
"data": {
"image/svg+xml": [
""
]
},
"metadata": {},
"output_type": "display_data"
},
{
"data": {
"image/svg+xml": [
""
]
},
"metadata": {},
"output_type": "display_data"
},
{
"data": {
"image/svg+xml": [
""
]
},
"metadata": {},
"output_type": "display_data"
},
{
"data": {
"image/svg+xml": [
""
]
},
"metadata": {},
"output_type": "display_data"
}
],
"source": [
"smilesNAG = ['CC(=O)N[C@H]1[C@H](O)O[C@H](CO)[C@@H](O)[C@@H]1O']\n",
"all_smiles = ['CC(=O)N[C@H]1[C@H](O)O[C@H](CO)[C@@H](O)[C@@H]1O', # original\n",
" 'OC(=O)N[C@H]1[C@H](O)O[C@H](CO)[C@@H](O)[C@@H]1O', # OH \n",
" '[NH3+]C(=O)N[C@H]1[C@H](O)O[C@H](CO)[C@@H](O)[C@@H]1O', # NH3+\n",
" 'C(=O)N[C@H]1[C@H](O)O[C@H](CO)[C@@H](O)[C@@H]1O', # H\n",
" 'C2=CC=C(C=C2)C(=O)N[C@H]1[C@H](O)O[C@H](CO)[C@@H](O)[C@@H]1O', # Ph\n",
" '[O-]C(=O)C(=O)N[C@H]1[C@H](O)O[C@H](CO)[C@@H](O)[C@@H]1O'] # COO-\n",
"mols= []\n",
"images =[]\n",
"\n",
"for s in all_smiles:\n",
" m = oddt.toolkit.readstring('smi', s)\n",
" if not m.OBMol.Has3D(): \n",
" m.make3D(forcefield='mmff94', steps=150)\n",
" m.removeh()\n",
" m.OBMol.AddPolarHydrogens()\n",
"\n",
" mols.append(m)\n",
" ###with print m.OBMol.Has3D() was found that:\n",
" ### deep copy needed to keep 3D , write svg make mols flat\n",
" images.append((SVG(copy.deepcopy(m).write('svg'))))\n",
"\n",
"display_svg(*images)"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"### Докинг"
]
},
{
"cell_type": "code",
"execution_count": 8,
"metadata": {},
"outputs": [
{
"name": "stdout",
"output_type": "stream",
"text": [
"/tmp/autodock_vina_cxEUSJ\n"
]
},
{
"data": {
"text/html": [
"Параметры докинга"
],
"text/plain": [
""
]
},
"metadata": {},
"output_type": "display_data"
},
{
"name": "stdout",
"output_type": "stream",
"text": [
"Координаты центра\n",
"--center_x 40.7385450082 --center_y 30.4164680851 --center_z 22.7187446809\n",
"Количество точек по каждой оси\n",
"--size_x 20 --size_y 20 --size_z 20\n",
"Ускорители\n",
"--cpu 1\n",
"Насколько глубоко ведется поиск (полнота алгоритма)\n",
"--exhaustiveness 8\n",
"Наибольшее число объектов связываний, которое может быть сгенерировано\n",
"--num_modes 3\n",
"Максимально допустимый разброс энергии между ллучшим и худшим вариантом связывания\n",
"--energy_range 3\n"
]
}
],
"source": [
"dock_obj= oddt.docking.AutodockVina.autodock_vina(\n",
" protein=prot,size=(20,20,20),center=center,\n",
" executable='/usr/bin/vina',autocleanup=True, num_modes=3)\n",
"\n",
"print dock_obj.tmp_dir\n",
"#print \" \".join(dock_obj.params) # Опишите выдачу\n",
"display(HTML('Параметры докинга'))\n",
"print 'Координаты центра'\n",
"print \" \".join(dock_obj.params[:6]) \n",
"print 'Количество точек по каждой оси'\n",
"print \" \".join(dock_obj.params[6:12]) \n",
"print 'Ускорители'\n",
"print \" \".join(dock_obj.params[12:14]) \n",
"print 'Насколько глубоко ведется поиск (полнота алгоритма)'\n",
"print \" \".join(dock_obj.params[14:16]) \n",
"print 'Наибольшее число объектов связываний, которое может быть сгенерировано'\n",
"print \" \".join(dock_obj.params[16:18]) \n",
"print 'Максимально допустимый разброс энергии между ллучшим и худшим вариантом связывания'\n",
"print \" \".join(dock_obj.params[18:]) "
]
},
{
"cell_type": "code",
"execution_count": 9,
"metadata": {
"collapsed": true
},
"outputs": [],
"source": [
"# do it\n",
"res = dock_obj.dock(mols,prot)"
]
},
{
"cell_type": "code",
"execution_count": 10,
"metadata": {
"collapsed": true
},
"outputs": [],
"source": [
"info = np.array([[i,r.formula, r.data['vina_affinity'], r.data['vina_rmsd_ub'], r.residues[0].name] for i,r in enumerate(res)])"
]
},
{
"cell_type": "code",
"execution_count": 22,
"metadata": {},
"outputs": [
{
"data": {
"text/html": [
"\n",
"\n",
"
\n",
" \n",
" \n",
" | \n",
" formula | \n",
" residue | \n",
" type | \n",
" vina_affinity | \n",
" vina_rmsd_ub | \n",
"
\n",
" \n",
" \n",
" \n",
" | 12 | \n",
" C13H17NO6 | \n",
" UNL | \n",
" Ph | \n",
" -6.2 | \n",
" 0.000 | \n",
"
\n",
" \n",
" | 15 | \n",
" C8H13NO8 | \n",
" UNL | \n",
" COO- | \n",
" -6.0 | \n",
" 0.000 | \n",
"
\n",
" \n",
" | 13 | \n",
" C13H17NO6 | \n",
" UNL | \n",
" Ph | \n",
" -6.0 | \n",
" 6.706 | \n",
"
\n",
" \n",
" | 0 | \n",
" C8H15NO6 | \n",
" UNL | \n",
" original | \n",
" -5.9 | \n",
" 0.000 | \n",
"
\n",
" \n",
" | 6 | \n",
" C7H15N2O6 | \n",
" UNL | \n",
" NH3+ | \n",
" -5.7 | \n",
" 0.000 | \n",
"
\n",
" \n",
" | 9 | \n",
" C7H13NO6 | \n",
" UNL | \n",
" H | \n",
" -5.7 | \n",
" 0.000 | \n",
"
\n",
" \n",
" | 3 | \n",
" C7H13NO7 | \n",
" UNL | \n",
" OH | \n",
" -5.7 | \n",
" 0.000 | \n",
"
\n",
" \n",
" | 10 | \n",
" C7H13NO6 | \n",
" UNL | \n",
" H | \n",
" -5.3 | \n",
" 2.777 | \n",
"
\n",
" \n",
" | 7 | \n",
" C7H15N2O6 | \n",
" UNL | \n",
" NH3+ | \n",
" -5.2 | \n",
" 5.096 | \n",
"
\n",
" \n",
" | 11 | \n",
" C7H13NO6 | \n",
" UNL | \n",
" H | \n",
" -5.2 | \n",
" 2.782 | \n",
"
\n",
" \n",
" | 4 | \n",
" C7H13NO7 | \n",
" UNL | \n",
" OH | \n",
" -5.2 | \n",
" 3.274 | \n",
"
\n",
" \n",
" | 14 | \n",
" C13H17NO6 | \n",
" UNL | \n",
" Ph | \n",
" -5.2 | \n",
" 2.323 | \n",
"
\n",
" \n",
" | 16 | \n",
" C8H13NO8 | \n",
" UNL | \n",
" COO- | \n",
" -5.2 | \n",
" 5.614 | \n",
"
\n",
" \n",
" | 17 | \n",
" C8H13NO8 | \n",
" UNL | \n",
" COO- | \n",
" -5.2 | \n",
" 4.271 | \n",
"
\n",
" \n",
" | 5 | \n",
" C7H13NO7 | \n",
" UNL | \n",
" OH | \n",
" -5.1 | \n",
" 3.479 | \n",
"
\n",
" \n",
" | 1 | \n",
" C8H15NO6 | \n",
" UNL | \n",
" original | \n",
" -5.1 | \n",
" 3.237 | \n",
"
\n",
" \n",
" | 8 | \n",
" C7H15N2O6 | \n",
" UNL | \n",
" NH3+ | \n",
" -5.0 | \n",
" 5.167 | \n",
"
\n",
" \n",
" | 2 | \n",
" C8H15NO6 | \n",
" UNL | \n",
" original | \n",
" -5.0 | \n",
" 5.143 | \n",
"
\n",
" \n",
"
\n",
"