{
 "cells": [
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# How to load new libraries to Chemical Checker infrastructure"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "\n",
    "## Load library files to datasource table\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "import os\n",
    "import sys\n",
    "sys.path.append(\"/aloy/home/oguitart/projects/source/chemical_checker/chemicalchecker\")\n",
    "\n",
    "os.environ[\"CC_CONFIG\"] = '/aloy/home/oguitart/projects/source/chemical_checker/chemicalchecker/tests/data/config.json'\n",
    "\n",
    "\n",
    "from chemicalchecker.util import Config\n",
    "\n",
    "from chemicalchecker.database import Datasource\n",
    "\n",
    "\n",
    "\n",
    "biur_real = {'name':\"biur_real\",'url':\"file:///aloy/scratch/sbnb-adm/checker/local_downloads/BIUR_20100326.sdf\",\n",
    "             'permanent':True,'enabled':False,'user':\"\",'password':\"\",'description':\"ChemistryX Biur real library\",\n",
    "             'molrepo_name':\"biur_real\",'molrepo_file':\"BIUR_20100326.sdf\",'is_db':False}\n",
    "biur_virtual = {'name':\"biur_virtual\",'url':\"file:///aloy/scratch/sbnb-adm/checker/local_downloads/VIRTUAL_BIUR_POR_MW.tar.gz\",\n",
    "                'permanent':True,'enabled':False,'user':\"\",'password':\"\",'description':\"ChemistryX Biur virtual library\",\n",
    "                'molrepo_name':\"biur_virtual\",'molrepo_file':\"VIRTUAL_BIUR_POR_MW\",'is_db':False}\n",
    "\n",
    "\n",
    "Datasource.add(biur_real)\n",
    "Datasource.add(biur_virtual)\n"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Add Library parser\n",
    "\n",
    "Check the *package/chemicalchecker/util/parser/parser.py* file and add a method for each one of the new libraries to parse those libraries. \n",
    "The new method to parse the library should have the same name as the **molrepo_name** used in the datasource table. The output of the method should be the same as the other methods used in the **parser.py** file."
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Download  & parse datasource associated to library \"biur_real\"\n",
    "\n",
    "1. Download the datasource files\n",
    "2. Parse downloaded files and load it to the molrepo table"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "scrolled": true
   },
   "outputs": [],
   "source": [
    "import os\n",
    "import sys\n",
    "sys.path.append(\"/aloy/home/oguitart/projects/source/chemical_checker/chemicalchecker\")\n",
    "\n",
    "os.environ[\"CC_CONFIG\"] = '/aloy/home/oguitart/projects/source/chemical_checker/chemicalchecker/tests/data/config.json'\n",
    "\n",
    "\n",
    "from chemicalchecker.util import Config\n",
    "\n",
    "from chemicalchecker.database import Datasource, Molrepo\n",
    "\n",
    "# This method downloads all datasources related to this library \n",
    "# and parse the files to load the new data to the molrepo table\n",
    "\n",
    "Molrepo.from_molrepo_name(\"biur_real\")\n"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Download datasource associated to library \"biur_virtual\"\n",
    "\n",
    "1. Download the datasource files\n",
    "2. Parse downloaded files and load it to the molrepo table"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "scrolled": true
   },
   "outputs": [],
   "source": [
    "import os\n",
    "import sys\n",
    "sys.path.append(\"/aloy/home/oguitart/projects/source/chemical_checker/chemicalchecker\")\n",
    "\n",
    "os.environ[\"CC_CONFIG\"] = '/aloy/home/oguitart/projects/source/chemical_checker/chemicalchecker/tests/data/config.json'\n",
    "\n",
    "\n",
    "from chemicalchecker.util import Config\n",
    "\n",
    "from chemicalchecker.database import Datasource, Molrepo\n",
    "\n",
    "# This method downloads all datasources related to this library \n",
    "# and parse the files to load the new data to the molrepo table\n",
    "\n",
    "Molrepo.from_molrepo_name(\"biur_virtual\")\n"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Calculate the molecular properties and store them for library \"biur_real\"\n",
    "\n",
    "It will also add the missing molecules to the structure table\n",
    "It will create jobs to do this task in the cluster.  \n",
    "It waits for a molecular property calculation to finish before submitting another one. \n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "scrolled": true
   },
   "outputs": [],
   "source": [
    "import os\n",
    "import sys\n",
    "import shutil\n",
    "\n",
    "sys.path.append(\"/aloy/home/oguitart/projects/source/chemical_checker/chemicalchecker\")\n",
    "\n",
    "os.environ[\"CC_CONFIG\"] = '/aloy/home/oguitart/projects/source/chemical_checker/chemicalchecker/tests/data/config.json'\n",
    "\n",
    "\n",
    "from chemicalchecker.database import Molprop, Structure\n",
    "from chemicalchecker.database import Molrepo\n",
    "from chemicalchecker.util import HPC\n",
    "from chemicalchecker.util import Config\n",
    "\n",
    "\n",
    "molprops = ['fp2d','fp3d','subskeys','scaffolds','physchem']\n",
    "\n",
    "molrepo_ik_inchi = Molrepo.get_fields_by_molrepo_name(\"biur_real\",[\"inchikey\",\"inchi\"])\n",
    "\n",
    "for mol in molprops:\n",
    "\n",
    "    job_path = \"/aloy/scratch/oguitart/jobs_molprop_\" + mol\n",
    "    if os.path.isdir(job_path):\n",
    "        shutil.rmtree(job_path)\n",
    "    os.mkdir(job_path)\n",
    "\n",
    "    molprop = Molprop(mol)\n",
    "\n",
    "    # This method sends the job and waits for the job to finish\n",
    "    cluster = molprop.molprop_hpc(job_path,molrepo_ik_inchi)\n",
    "   \n"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Calculate the molecular properties and store them for library \"biur_virtual\"\n",
    "\n",
    "It will create jobs to do this task in the cluster.  \n",
    "It waits for a molecular property calculation to finish before submitting another one."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "scrolled": true
   },
   "outputs": [],
   "source": [
    "import os\n",
    "import sys\n",
    "import shutil\n",
    "\n",
    "sys.path.append(\"/aloy/home/oguitart/projects/source/chemical_checker/chemicalchecker\")\n",
    "\n",
    "os.environ[\"CC_CONFIG\"] = '/aloy/home/oguitart/projects/source/chemical_checker/chemicalchecker/tests/data/config.json'\n",
    "\n",
    "cc_config = '/aloy/home/oguitart/projects/source/chemical_checker/chemicalchecker/tests/data/config.json'\n",
    "\n",
    "from chemicalchecker.database import Molprop, Structure\n",
    "from chemicalchecker.database import Molrepo\n",
    "from chemicalchecker.util import HPC\n",
    "from chemicalchecker.util import Config\n",
    "\n",
    "\n",
    "molprops = ['fp2d','fp3d','subskeys','scaffolds','physchem']\n",
    "\n",
    "molrepo_ik_inchi = Molrepo.get_fields_by_molrepo_name(\"biur_virtual\",[\"inchikey\",\"inchi\"])\n",
    "\n",
    "for mol in molprops:\n",
    "\n",
    "    job_path = \"/aloy/scratch/oguitart/jobs_molprop_\" + mol\n",
    "    if os.path.isdir(job_path):\n",
    "        shutil.rmtree(job_path)\n",
    "    os.mkdir(job_path)\n",
    "\n",
    "    molprop = Molprop(mol)\n",
    "\n",
    "    # This method sends the job and waits for the job to finish\n",
    "    cluster = molprop.molprop_hpc(job_path,molrepo_ik_inchi)\n",
    "    \n"
   ]
  }
 ],
 "metadata": {
  "kernelspec": {
   "display_name": "Python 2",
   "language": "python",
   "name": "python2"
  },
  "language_info": {
   "codemirror_mode": {
    "name": "ipython",
    "version": 2
   },
   "file_extension": ".py",
   "mimetype": "text/x-python",
   "name": "python",
   "nbconvert_exporter": "python",
   "pygments_lexer": "ipython2",
   "version": "2.7.10"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 2
}