5-Knowledge Graph-AAC.ipynb

{
 "cells": [
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "#! pip install neo4j"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# Creating a knowledge graph in neo4j \n",
    "\n",
    "Here we provide a sample notebook for KG construction with hypothetical graph data where there are 3 typrs of nodes and one relationship. Following are the general steps for KG set up for personal PC:"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "- First download and install the neo4j Desktop on your computer.\n",
    "- Create a new project.\n",
    "- Click on ```ADD``` tab right to the project and add local DBMS. Make sure your user and passwords are noted for future access.\n",
    "- Install ```apoc``` library through plugin.\n",
    "- Open config folder and put provided ```neo4j.config``` and ```apoc.config```.\n",
    "    \ta. Click on the 3 dots on the right of the graph DBMS.\n",
    "\t\tb. Click open\n",
    "\t\tc. Click configuration\n",
    "\t\td. Copy and paste provided neo4j.config and apoc.config there.\n",
    "- Start the graph database\n",
    "- Click open tab of the DBMS to open neo4j Browser.\n",
    "- Install ```anaconda``` python and run ```Jupyter application```\n",
    "- Use this notebook to communicate with graph database"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "metadata": {},
   "outputs": [],
   "source": [
    "import pandas as pd\n",
    "import json\n",
    "import time\n",
    "from neo4j import GraphDatabase"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "#### Make a connection with Graph Database"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 1,
   "metadata": {},
   "outputs": [],
   "source": [
    "import pandas as pd\n",
    "from neo4j import GraphDatabase\n",
    "driver=GraphDatabase.driver(uri=\"bolt://127.0.0.1:7687\",auth=(\"neo4j\",\"12345678\"))\n",
    "session=driver.session()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "<neo4j._sync.driver.BoltDriver at 0x7fc41b459d90>"
      ]
     },
     "execution_count": 3,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "driver"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "#### Create Constraints for uniqueness of nodes"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 58,
   "metadata": {},
   "outputs": [],
   "source": [
    "def create_constraints(driver):\n",
    "        '''\n",
    "        Create constraints for nodes in the graph\n",
    "        args:\n",
    "            driver: neo4j driver\n",
    "        returns:\n",
    "            None\n",
    "        '''\n",
    "        query = [\"CREATE CONSTRAINT UniqueOrganismNameConstraint FOR (O:Organism) REQUIRE O.name IS UNIQUE\",\\\n",
    "                \"CREATE CONSTRAINT UniqueProteinIdConstraint FOR (P:Protein) REQUIRE P.UID IS UNIQUE\",\\\n",
    "                \"CREATE CONSTRAINT UniquePTMTypeNameConstraint FOR (PTM:PTM_Type) REQUIRE PTM.name IS UNIQUE\",\\\n",
    "                \"CREATE CONSTRAINT UniqueCellLineNameConstraint FOR (C:cellLine) REQUIRE C.name IS UNIQUE\",\\\n",
    "                \"CREATE CONSTRAINT UniqueReferenceIdConstraint FOR (R:references) REQUIRE R.Pmid IS UNIQUE\",\\\n",
    "                \"CREATE CONSTRAINT UniquePathwayIdConstraint FOR (Pw: Pathway) REQUIRE Pw.ID IS UNIQUE\"]\n",
    "        \n",
    "        query2 = [ \"CREATE CONSTRAINT UniqueLocationConstraint FOR (L:Location) REQUIRE L.LID IS UNIQUE\"]\n",
    "\n",
    "        with driver.session() as session:\n",
    "            for constraint in query2:\n",
    "                session.run(constraint)\n",
    "\n",
    "\n",
    "                "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 70,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "\n",
      "        Create constraints for nodes in the graph\n",
      "        args:\n",
      "            driver: neo4j driver\n",
      "        returns:\n",
      "            None\n",
      "        \n"
     ]
    }
   ],
   "source": [
    "print(create_constraints.__doc__)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 60,
   "metadata": {},
   "outputs": [],
   "source": [
    "'''UNCOMMENT AND RUN THIS CELL ONLY ONCE'''\n",
    "#create_constraints(driver)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Populate Node\n",
    "- There are three arbitrary node types\n",
    "- There are one edge types between nodes."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 71,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "{'Organism': 'Mus musculus',\n",
       " 'Uniprot_AC': 'Q61598',\n",
       " 'position': 317,\n",
       " 'standard_seq': 'EKVGQVIRVICILSHPIKNTNDANSCQIIIPQNQVNRKSDIYVCMISFAHN',\n",
       " 'PTM_Catalogy': 'Oxidation PTM',\n",
       " 'PTM_Type': 'S-nitrosylation',\n",
       " 'ref': '28710281',\n",
       " 'cell_line': 'Ischemia heart cells',\n",
       " 'Identification_Strategy': 'Thiol isotope labeling',\n",
       " 'Identification_Technique': 'SNOxICAT'}"
      ]
     },
     "execution_count": 71,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "{'Organism': 'Mus musculus',\n",
    " 'Uniprot_AC': 'Q61598',\n",
    " 'position': 317,\n",
    " 'standard_seq': 'EKVGQVIRVICILSHPIKNTNDANSCQIIIPQNQVNRKSDIYVCMISFAHN',\n",
    " 'PTM_Catalogy': 'Oxidation PTM',\n",
    " 'PTM_Type': 'S-nitrosylation',\n",
    " 'ref': '28710281',\n",
    " 'cell_line': 'Ischemia heart cells',\n",
    " 'Identification_Strategy': 'Thiol isotope labeling',\n",
    " 'Identification_Technique': 'SNOxICAT'}"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 72,
   "metadata": {},
   "outputs": [],
   "source": [
    "def create_organism_node(data):\n",
    "        '''  \n",
    "        Create Organism node in the graph\n",
    "         args:\n",
    "            data: dictionary containing the data for the node\n",
    "        returns:\n",
    "            None\n",
    "        '''\n",
    "        def tx_function(tx,data):\n",
    "            query = \"WITH '\" + data  + \"' as url \\\n",
    "            CALL apoc.load.json(url) YIELD value \\\n",
    "            MERGE (O:Organism{name:value.Organism})\"\n",
    "            \n",
    "            #print(query)\n",
    "            tx.run(query,data=data)\n",
    "        \n",
    "        with driver.session() as session:   \n",
    "            session.execute_write(tx_function,data)\n",
    "            \n",
    "\n",
    "            \n",
    "def create_protein_node(data):\n",
    "        '''\n",
    "        Create Protein node in the graph\n",
    "        args:\n",
    "            data: dictionary containing the data for the node\n",
    "        returns:\n",
    "            None\n",
    "        '''\n",
    "        def tx_function(tx,data):\n",
    "            query = \"WITH '\" + data  + \"' as url \\\n",
    "            CALL apoc.load.json(url) YIELD value \\\n",
    "            MERGE (P:Protein{UID:value.Uniprot_AC})\\\n",
    "            ON CREATE SET P.standard_seq=value.standard_seq,\\\n",
    "            P.position=value.position\"\n",
    "            \n",
    "            #print(query)\n",
    "            tx.run(query,data=data)\n",
    "        \n",
    "        with driver.session() as session:   \n",
    "            session.execute_write(tx_function,data)\n",
    "\n",
    "\n",
    "\n",
    "            \n",
    "def create_ptmtype_node(data):\n",
    "        '''\n",
    "        Create PTM_Type node in the graph\n",
    "        args:\n",
    "            data: dictionary containing the data for the node\n",
    "        returns:\n",
    "            None\n",
    "        '''\n",
    "        def tx_function(tx,data):\n",
    "            query = \"WITH '\" + data  + \"' as url \\\n",
    "            CALL apoc.load.json(url) YIELD value \\\n",
    "            MERGE (PTM:PTM_Type{name:value.PTM_Type})\\\n",
    "            ON CREATE SET PTM.PTM_Catalogy=value.PTM_Catalogy\"\n",
    "            \n",
    "            #print(query)\n",
    "            tx.run(query,data=data)\n",
    "        with driver.session() as session:   \n",
    "            session.execute_write(tx_function,data)\n",
    "\n",
    "\n",
    "            \n",
    "\n",
    "def create_cellline_node(data):\n",
    "        '''\n",
    "        Create cellLine node in the graph\n",
    "        args:\n",
    "            data: dictionary containing the data for the node\n",
    "        returns:\n",
    "            None\n",
    "        '''\n",
    "        def tx_function(tx,data):\n",
    "            query = \"WITH '\" + data  + \"' as url \\\n",
    "            CALL apoc.load.json(url) YIELD value \\\n",
    "            MERGE (C:cellLine{name:value.cell_line})\\\n",
    "            ON CREATE SET C.Identification_Strategy=value.Identification_Strategy,\\\n",
    "            C.Identification_Technique=value.Identification_Technique\"\n",
    "            \n",
    "            #print(query)\n",
    "            tx.run(query,data=data)\n",
    "        with driver.session() as session:   \n",
    "            session.execute_write(tx_function,data)\n",
    "\n",
    "\n",
    "\n",
    "\n",
    "\n",
    "def create_reference_node(data):\n",
    "        '''\n",
    "        Create references node in the graph\n",
    "        args:\n",
    "            data: dictionary containing the data for the node\n",
    "        returns:\n",
    "            None\n",
    "        '''\n",
    "        def tx_function(tx,data):\n",
    "            query = \"WITH '\" + data  + \"' as url \\\n",
    "            CALL apoc.load.json(url) YIELD value \\\n",
    "            MERGE (R:references{Pmid:value.ref})\"\n",
    "            \n",
    "            \n",
    "            #print(query)\n",
    "            tx.run(query,data=data)\n",
    "        \n",
    "        with driver.session() as session:   \n",
    "            session.execute_write(tx_function,data)\n",
    "\n",
    "\n",
    "def create_location_node(data):\n",
    "        '''\n",
    "        Create references node in the graph\n",
    "        args:\n",
    "            data: dictionary containing the data for the node\n",
    "        returns:\n",
    "            None\n",
    "        '''\n",
    "        def tx_function(tx,data):\n",
    "            query = \"WITH '\" + data  + \"' as url \\\n",
    "            CALL apoc.load.json(url) YIELD value \\\n",
    "            MERGE (L:Location{LID:value.LID})\\\n",
    "            ON CREATE SET L.Sequence=value.standard_seq,\\\n",
    "            L.Location=value.position\"\n",
    "            \n",
    "            \n",
    "            #print(query)\n",
    "            tx.run(query,data=data)\n",
    "        \n",
    "        with driver.session() as session:   \n",
    "            session.execute_write(tx_function,data)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 73,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "success! total time:  0.29370999336242676\n"
     ]
    }
   ],
   "source": [
    "t1 = time.time()\n",
    "data = \"data.json\"\n",
    "create_organism_node(data)\n",
    "t2 = time.time()\n",
    "print( \"success! total time: \", t2-t1)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 74,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "success! total time:  0.5182251930236816\n"
     ]
    }
   ],
   "source": [
    "t1 = time.time()\n",
    "data = \"data.json\"\n",
    "create_protein_node(data)\n",
    "t2 = time.time()\n",
    "print( \"success! total time: \", t2-t1)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 75,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "success! total time:  0.23247933387756348\n"
     ]
    }
   ],
   "source": [
    "t1 = time.time()\n",
    "data = \"data.json\"\n",
    "create_ptmtype_node(data)\n",
    "t2 = time.time()\n",
    "print( \"success! total time: \", t2-t1)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 76,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "success! total time:  0.23611187934875488\n"
     ]
    }
   ],
   "source": [
    "t1 = time.time()\n",
    "data = \"data.json\"\n",
    "create_cellline_node(data)\n",
    "t2 = time.time()\n",
    "print( \"success! total time: \", t2-t1)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 77,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "success! total time:  0.19608235359191895\n"
     ]
    }
   ],
   "source": [
    "t1 = time.time()\n",
    "data = \"data.json\"\n",
    "create_reference_node(data)\n",
    "t2 = time.time()\n",
    "print( \"success! total time: \", t2-t1)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 78,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "success! total time:  0.8331589698791504\n"
     ]
    }
   ],
   "source": [
    "t1 = time.time()\n",
    "data = \"data.json\"\n",
    "create_location_node(data)\n",
    "t2 = time.time()\n",
    "print( \"success! total time: \", t2-t1)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Create Edges"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 81,
   "metadata": {},
   "outputs": [],
   "source": [
    "def create_edge_protein2location(data):\n",
    "        def tx_function(tx,data):\n",
    "            query = \"WITH '\" + data  + \"' as url \\\n",
    "            CALL apoc.load.json(url) YIELD value \\\n",
    "            MATCH (P {UID:value.Uniprot_AC})\\\n",
    "            MATCH (L {LID:value.LID})\\\n",
    "            MERGE (P)-[:HAS_CYSTEINE]->(L)\"\n",
    "\n",
    "            tx.run(query,data=data)\n",
    "        with driver.session() as session:   \n",
    "            session.execute_write(tx_function,data)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 82,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "success! total time:  1126.3054032325745\n"
     ]
    }
   ],
   "source": [
    "t1 = time.time()\n",
    "data = \"data.json\"\n",
    "create_edge_protein2location(data)\n",
    "t2 = time.time()\n",
    "print( \"success! total time: \", t2-t1)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "metadata": {},
   "outputs": [],
   "source": [
    "def create_edge_location2ptm_type(data):\n",
    "        def tx_function(tx,data):\n",
    "            query = \"WITH '\" + data  + \"' as url \\\n",
    "            CALL apoc.load.json(url) YIELD value \\\n",
    "            MATCH (L {LID:value.LID})\\\n",
    "            MATCH (PTM {name:value.PTM_Type})\\\n",
    "            MERGE (L)-[:HAS_PTM_TYPE]->(PTM)\"\n",
    "\n",
    "            tx.run(query,data=data)\n",
    "        with driver.session() as session:   \n",
    "            session.execute_write(tx_function,data)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "success! total time:  1044.7420127391815\n"
     ]
    }
   ],
   "source": [
    "t1=time.time()\n",
    "data = \"data.json\"\n",
    "create_edge_location2ptm_type(data)\n",
    "t2=time.time()\n",
    "print( \"success! total time: \", t2-t1)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 6,
   "metadata": {},
   "outputs": [],
   "source": [
    "def create_edge_ptm2cellline(data):\n",
    "        def tx_function(tx,data):\n",
    "            query = \"WITH '\" + data  + \"' as url \\\n",
    "            CALL apoc.load.json(url) YIELD value \\\n",
    "            MATCH (PTM {name:value.PTM_Type})\\\n",
    "            MATCH (C {name:value.cell_line})\\\n",
    "            MERGE (PTM)-[:FROM]->(C)\"\n",
    "\n",
    "            tx.run(query,data=data)\n",
    "        with driver.session() as session:   \n",
    "            session.execute_write(tx_function,data)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 7,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "success! total time:  845.1974122524261\n"
     ]
    }
   ],
   "source": [
    "t1=time.time()\n",
    "data = \"data.json\"\n",
    "create_edge_ptm2cellline(data)\n",
    "t2=time.time()\n",
    "print( \"success! total time: \", t2-t1)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 8,
   "metadata": {},
   "outputs": [],
   "source": [
    "def create_edge_cellline2reference(data):\n",
    "        def tx_function(tx,data):\n",
    "            query = \"WITH '\" + data  + \"' as url \\\n",
    "            CALL apoc.load.json(url) YIELD value \\\n",
    "            MATCH (C {name:value.cell_line})\\\n",
    "            MATCH (R {Pmid:value.ref})\\\n",
    "            MERGE (C)-[:HAS_REFERENCE]->(R)\"\n",
    "\n",
    "            tx.run(query,data=data)\n",
    "        with driver.session() as session:   \n",
    "            session.execute_write(tx_function,data)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 9,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "success! total time:  829.3450391292572\n"
     ]
    }
   ],
   "source": [
    "t1=time.time()\n",
    "data = \"data.json\"\n",
    "create_edge_cellline2reference(data)\n",
    "t2=time.time()\n",
    "print( \"success! total time: \", t2-t1)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 10,
   "metadata": {},
   "outputs": [],
   "source": [
    "def create_edge_ptm2reference(data):\n",
    "        def tx_function(tx,data):\n",
    "            query = \"WITH '\" + data  + \"' as url \\\n",
    "            CALL apoc.load.json(url) YIELD value \\\n",
    "            MATCH (PTM {name:value.PTM_Type})\\\n",
    "            MATCH (R {Pmid:value.ref})\\\n",
    "            MERGE (PTM)-[:HAS_REFERENCE]->(R)\"\n",
    "\n",
    "            tx.run(query,data=data)\n",
    "        with driver.session() as session:   \n",
    "            session.execute_write(tx_function,data)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 11,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "success! total time:  828.799971818924\n"
     ]
    }
   ],
   "source": [
    "t1=time.time()\n",
    "data = \"data.json\"\n",
    "create_edge_ptm2reference(data)\n",
    "t2=time.time()\n",
    "print( \"success! total time: \", t2-t1)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 12,
   "metadata": {},
   "outputs": [],
   "source": [
    "def create_edge_ptm2organism(data):\n",
    "        def tx_function(tx,data):\n",
    "            query = \"WITH '\" + data  + \"' as url \\\n",
    "            CALL apoc.load.json(url) YIELD value \\\n",
    "            MATCH (PTM {name:value.PTM_Type})\\\n",
    "            MATCH (O {name:value.Organism})\\\n",
    "            MERGE (PTM)-[:BELONGS_TO]->(O)\"\n",
    "\n",
    "            tx.run(query,data=data)\n",
    "        with driver.session() as session:   \n",
    "            session.execute_write(tx_function,data)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 13,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "success! total time:  848.8454530239105\n"
     ]
    }
   ],
   "source": [
    "t1=time.time()\n",
    "data = \"data.json\"\n",
    "create_edge_ptm2organism(data)\n",
    "t2=time.time()\n",
    "print( \"success! total time: \", t2-t1)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 14,
   "metadata": {},
   "outputs": [],
   "source": [
    "def create_edge_protein2organism(data):\n",
    "        def tx_function(tx,data):\n",
    "            query = \"WITH '\" + data  + \"' as url \\\n",
    "            CALL apoc.load.json(url) YIELD value \\\n",
    "            MATCH (P {UID:value.Uniprot_AC})\\\n",
    "            MATCH (O {name:value.Organism})\\\n",
    "            MERGE (P)-[:BELONGS_TO]->(O)\"\n",
    "        \n",
    "\n",
    "            tx.run(query,data=data)\n",
    "        with driver.session() as session:   \n",
    "            session.execute_write(tx_function,data)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 15,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "success! total time:  856.0041120052338\n"
     ]
    }
   ],
   "source": [
    "t1=time.time()\n",
    "data = \"data.json\"\n",
    "create_edge_protein2organism(data)\n",
    "t2=time.time()\n",
    "print( \"success! total time: \", t2-t1)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Reactome Update"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 41,
   "metadata": {},
   "outputs": [],
   "source": [
    "def create_pathway_node_small(data):\n",
    "        '''  \n",
    "        Create Pathway node in the graph\n",
    "         args:\n",
    "            data: dictionary containing the data for the node\n",
    "        returns:\n",
    "            None\n",
    "        '''\n",
    "        def tx_function(tx,data):\n",
    "            query = \"WITH '\" + data  + \"' as url \\\n",
    "            CALL apoc.load.json(url) YIELD value \\\n",
    "            MERGE (Pw:Pathway{ID:value.Pathway_identifier})\\\n",
    "            ON CREATE SET Pw.name=value.Pathway_name\"\n",
    "            \n",
    "            #print(query)\n",
    "            tx.run(query,data=data)\n",
    "        \n",
    "        with driver.session() as session:   \n",
    "            session.execute_write(tx_function,data)\n",
    "\n",
    "def create_pathway_node_large(data):\n",
    "        '''  \n",
    "        Create Pathway node in the graph\n",
    "         args:\n",
    "            data: dictionary containing the data for the node\n",
    "        returns:\n",
    "            None\n",
    "        '''\n",
    "        def tx_function(tx,data):\n",
    "            query = \"WITH '\" + data  + \"' as url \\\n",
    "            CALL apoc.load.json(url) YIELD value \\\n",
    "            MERGE (Pw:Pathway{ID:value.ID})\\\n",
    "            ON CREATE SET Pw.name=value.name\"\n",
    "            \n",
    "            #print(query)\n",
    "            tx.run(query,data=data)\n",
    "        \n",
    "        with driver.session() as session:   \n",
    "            session.execute_write(tx_function,data)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 42,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "success! total time:  0.626776933670044\n"
     ]
    }
   ],
   "source": [
    "t1 = time.time()\n",
    "data = \"pathway-data-large.json\"\n",
    "create_pathway_node_large(data)\n",
    "t2 = time.time()\n",
    "print( \"success! total time: \", t2-t1)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 44,
   "metadata": {},
   "outputs": [],
   "source": [
    "def create_edge_protein2pathway_small(data):\n",
    "        def tx_function(tx,data):\n",
    "            query = \"WITH '\" + data  + \"' as url \\\n",
    "            CALL apoc.load.json(url) YIELD value \\\n",
    "            MATCH (P {UID:value.Mapped_entities})\\\n",
    "            MATCH (Pw {ID:value.Pathway_identifier})\\\n",
    "            MERGE (P)-[:ASSOCIATED]->(Pw)\"\n",
    "\n",
    "            tx.run(query,data=data)\n",
    "        with driver.session() as session:   \n",
    "            session.execute_write(tx_function,data)\n",
    "\n",
    "def create_edge_protein2pathway_large(data):\n",
    "        def tx_function(tx,data):\n",
    "            query = \"WITH '\" + data  + \"' as url \\\n",
    "            CALL apoc.load.json(url) YIELD value \\\n",
    "            MATCH (P {UID:value.protein})\\\n",
    "            MATCH (Pw {ID:value.ID})\\\n",
    "            MERGE (P)-[:ASSOCIATED]->(Pw)\"\n",
    "\n",
    "            tx.run(query,data=data)\n",
    "        with driver.session() as session:   \n",
    "            session.execute_write(tx_function,data)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 45,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "success! total time:  1380.0068852901459\n"
     ]
    }
   ],
   "source": [
    "t1 = time.time()\n",
    "data = \"pathway-data-large.json\"\n",
    "create_edge_protein2pathway_large(data)\n",
    "t2 = time.time()\n",
    "print( \"success! total time: \", t2-t1)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Add PPI data"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 28,
   "metadata": {},
   "outputs": [],
   "source": [
    "def create_ppi_proteinA_node(data):\n",
    "        '''  \n",
    "        Create PPI protein node in the graph\n",
    "         args:\n",
    "            data: dictionary containing the data for the node\n",
    "        returns:\n",
    "            None\n",
    "        '''\n",
    "        def tx_function(tx,data):\n",
    "            query = \"WITH '\" + data  + \"' as url \\\n",
    "            CALL apoc.load.json(url) YIELD value \\\n",
    "            MERGE (P:Protein{UID:value.IDA})\\\n",
    "            ON CREATE SET P.molecule=value.Molecule_A\"\n",
    "            \n",
    "            #print(query)\n",
    "            tx.run(query,data=data)\n",
    "        \n",
    "        with driver.session() as session:   \n",
    "            session.execute_write(tx_function,data)\n",
    "\n",
    "\n",
    "def create_ppi_proteinB_node(data):\n",
    "        '''  \n",
    "        Create PPI protein node in the graph\n",
    "         args:\n",
    "            data: dictionary containing the data for the node\n",
    "        returns:\n",
    "            None\n",
    "        '''\n",
    "        def tx_function(tx,data):\n",
    "            query = \"WITH '\" + data  + \"' as url \\\n",
    "            CALL apoc.load.json(url) YIELD value \\\n",
    "            MERGE (P:Protein{UID:value.IDA})\\\n",
    "            ON CREATE SET P.molecule=value.Molecule_B\"\n",
    "            \n",
    "            #print(query)\n",
    "            tx.run(query,data=data)\n",
    "        \n",
    "        with driver.session() as session:   \n",
    "            session.execute_write(tx_function,data)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": []
  },
  {
   "cell_type": "code",
   "execution_count": 29,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "success! total time:  0.06410813331604004\n"
     ]
    }
   ],
   "source": [
    "t1 = time.time()\n",
    "data = \"ppi.json\"\n",
    "create_ppi_proteinA_node(data)\n",
    "t2 = time.time()\n",
    "print( \"success! total time: \", t2-t1)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 30,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "success! total time:  0.016520977020263672\n"
     ]
    }
   ],
   "source": [
    "t1 = time.time()\n",
    "data = \"ppi.json\"\n",
    "create_ppi_proteinB_node(data)\n",
    "t2 = time.time()\n",
    "print( \"success! total time: \", t2-t1)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 31,
   "metadata": {},
   "outputs": [],
   "source": [
    "def create_edge_ppi(data):\n",
    "        def tx_function(tx,data):\n",
    "            query = \"WITH '\" + data  + \"' as url \\\n",
    "            CALL apoc.load.json(url) YIELD value \\\n",
    "            MATCH (PA:Protein {UID:value.IDA})\\\n",
    "            MATCH (PB:Protein {UID:value.IDB})\\\n",
    "            MERGE (PA)-[:HAS_PPI_INTERACTION]->(PB)\"\n",
    "\n",
    "            tx.run(query,data=data)\n",
    "        with driver.session() as session:   \n",
    "            session.execute_write(tx_function,data)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 32,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "success! total time:  0.03675699234008789\n"
     ]
    }
   ],
   "source": [
    "t1 = time.time()\n",
    "data = \"ppi.json\"\n",
    "create_edge_ppi(data)\n",
    "t2 = time.time()\n",
    "print( \"success! total time: \", t2-t1)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### References"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "https://neo4j.com/docs/cypher-cheat-sheet/current/"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Link Prediction"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 7,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "(15455, 1)"
      ]
     },
     "execution_count": 7,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "query = \"MATCH (P:Protein) RETURN P.UID\"\n",
    "\n",
    "res = []\n",
    "with driver.session() as session:\n",
    "    info = session.run(query)\n",
    "    for item in info:\n",
    "        res.append({\"ID\":item.values()[0]})\n",
    "df2 = pd.DataFrame(res)\n",
    "df2.shape"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 8,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "'Q61598'"
      ]
     },
     "execution_count": 8,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "df2['ID'][0]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 9,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "(13, 1)"
      ]
     },
     "execution_count": 9,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "query = \"MATCH (PTM:PTM_Type) RETURN PTM.name\"\n",
    "\n",
    "res = []\n",
    "with driver.session() as session:\n",
    "    info = session.run(query)\n",
    "    for item in info:\n",
    "        res.append({\"name\":item.values()[0]})\n",
    "df1 = pd.DataFrame(res)\n",
    "df1.shape"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 10,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "'S-nitrosylation'"
      ]
     },
     "execution_count": 10,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "df1['name'][0]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 11,
   "metadata": {},
   "outputs": [],
   "source": [
    "res = []\n",
    "\n",
    "for p in df2[\"ID\"]:\n",
    "    for ptm in df1[\"name\"]:\n",
    "    \n",
    "        query = \"MATCH (P:Protein {UID: '\" + p + \"'}) \\\n",
    "        MATCH (PTM:PTM_Type {name: '\" + ptm + \"'}) \\\n",
    "        RETURN gds.alpha.linkprediction.totalNeighbors(P, PTM, \\\n",
    "        {relationshipQuery: null})\\\n",
    "        AS Score, P.UID AS protein, PTM.name AS ptm_type\"\n",
    "\n",
    "        with driver.session() as session:\n",
    "            info = session.run(query)\n",
    "            for item in info:\n",
    "                res.append({\"Protein\":item.values()[1], \n",
    "                            \"PTM\":item.values()[2],\n",
    "                            \"Score\":item.values()[0]})"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 12,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>Protein</th>\n",
       "      <th>PTM</th>\n",
       "      <th>Score</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>37114</th>\n",
       "      <td>P49327</td>\n",
       "      <td>S-oxidation(nec.)</td>\n",
       "      <td>8278.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>37283</th>\n",
       "      <td>P04406</td>\n",
       "      <td>S-oxidation(nec.)</td>\n",
       "      <td>8276.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>37348</th>\n",
       "      <td>P10809</td>\n",
       "      <td>S-oxidation(nec.)</td>\n",
       "      <td>8276.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>37335</th>\n",
       "      <td>P21333</td>\n",
       "      <td>S-oxidation(nec.)</td>\n",
       "      <td>8276.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>41092</th>\n",
       "      <td>P46782</td>\n",
       "      <td>S-oxidation(nec.)</td>\n",
       "      <td>8275.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>37218</th>\n",
       "      <td>P13639</td>\n",
       "      <td>S-oxidation(nec.)</td>\n",
       "      <td>8274.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>37309</th>\n",
       "      <td>P23528</td>\n",
       "      <td>S-oxidation(nec.)</td>\n",
       "      <td>8273.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>40884</th>\n",
       "      <td>Q15365</td>\n",
       "      <td>S-oxidation(nec.)</td>\n",
       "      <td>8273.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>37140</th>\n",
       "      <td>P26641</td>\n",
       "      <td>S-oxidation(nec.)</td>\n",
       "      <td>8273.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>36984</th>\n",
       "      <td>P61247</td>\n",
       "      <td>S-oxidation(nec.)</td>\n",
       "      <td>8273.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>36958</th>\n",
       "      <td>Q00839</td>\n",
       "      <td>S-oxidation(nec.)</td>\n",
       "      <td>8273.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>37088</th>\n",
       "      <td>P55060</td>\n",
       "      <td>S-oxidation(nec.)</td>\n",
       "      <td>8272.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>37322</th>\n",
       "      <td>P22314</td>\n",
       "      <td>S-oxidation(nec.)</td>\n",
       "      <td>8272.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>37660</th>\n",
       "      <td>Q00610</td>\n",
       "      <td>S-oxidation(nec.)</td>\n",
       "      <td>8272.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>37205</th>\n",
       "      <td>P13797</td>\n",
       "      <td>S-oxidation(nec.)</td>\n",
       "      <td>8272.0</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "      Protein                PTM   Score\n",
       "37114  P49327  S-oxidation(nec.)  8278.0\n",
       "37283  P04406  S-oxidation(nec.)  8276.0\n",
       "37348  P10809  S-oxidation(nec.)  8276.0\n",
       "37335  P21333  S-oxidation(nec.)  8276.0\n",
       "41092  P46782  S-oxidation(nec.)  8275.0\n",
       "37218  P13639  S-oxidation(nec.)  8274.0\n",
       "37309  P23528  S-oxidation(nec.)  8273.0\n",
       "40884  Q15365  S-oxidation(nec.)  8273.0\n",
       "37140  P26641  S-oxidation(nec.)  8273.0\n",
       "36984  P61247  S-oxidation(nec.)  8273.0\n",
       "36958  Q00839  S-oxidation(nec.)  8273.0\n",
       "37088  P55060  S-oxidation(nec.)  8272.0\n",
       "37322  P22314  S-oxidation(nec.)  8272.0\n",
       "37660  Q00610  S-oxidation(nec.)  8272.0\n",
       "37205  P13797  S-oxidation(nec.)  8272.0"
      ]
     },
     "execution_count": 12,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "df_res = pd.DataFrame(res)\n",
    "df_res = df_res.sort_values(\"Score\", ascending = False)\n",
    "df_res.head(15)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 13,
   "metadata": {},
   "outputs": [],
   "source": [
    "df_res.to_csv(\"res.csv\", index = False)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 32,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "(200915, 3)"
      ]
     },
     "execution_count": 32,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "df_res.shape"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 33,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>S-glutathionylation</th>\n",
       "      <th>S-oxidation(nec.)</th>\n",
       "      <th>S-succination</th>\n",
       "      <th>S-carbonylation</th>\n",
       "      <th>S-prenylation</th>\n",
       "      <th>S-palmitoylation</th>\n",
       "      <th>S-nitrosylation</th>\n",
       "      <th>S-sulfenylation</th>\n",
       "      <th>S-sulfonylation</th>\n",
       "      <th>S-sulfinylation</th>\n",
       "      <th>S-disulfidation</th>\n",
       "      <th>S-itaconation</th>\n",
       "      <th>S-persulfidation</th>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>Protein</th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>M0R2D3</th>\n",
       "      <td>2984.0</td>\n",
       "      <td>8251.0</td>\n",
       "      <td>882.0</td>\n",
       "      <td>448.0</td>\n",
       "      <td>77.0</td>\n",
       "      <td>3151.0</td>\n",
       "      <td>4634.0</td>\n",
       "      <td>1893.0</td>\n",
       "      <td>71.0</td>\n",
       "      <td>366.0</td>\n",
       "      <td>257.0</td>\n",
       "      <td>862.0</td>\n",
       "      <td>1480.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>O54824</th>\n",
       "      <td>2983.0</td>\n",
       "      <td>8248.0</td>\n",
       "      <td>882.0</td>\n",
       "      <td>449.0</td>\n",
       "      <td>78.0</td>\n",
       "      <td>3153.0</td>\n",
       "      <td>4635.0</td>\n",
       "      <td>1896.0</td>\n",
       "      <td>71.0</td>\n",
       "      <td>367.0</td>\n",
       "      <td>258.0</td>\n",
       "      <td>859.0</td>\n",
       "      <td>1480.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>B4E2Q0</th>\n",
       "      <td>2981.0</td>\n",
       "      <td>8248.0</td>\n",
       "      <td>879.0</td>\n",
       "      <td>445.0</td>\n",
       "      <td>74.0</td>\n",
       "      <td>3148.0</td>\n",
       "      <td>4631.0</td>\n",
       "      <td>1892.0</td>\n",
       "      <td>68.0</td>\n",
       "      <td>363.0</td>\n",
       "      <td>254.0</td>\n",
       "      <td>859.0</td>\n",
       "      <td>1477.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>Q60649</th>\n",
       "      <td>2983.0</td>\n",
       "      <td>8248.0</td>\n",
       "      <td>882.0</td>\n",
       "      <td>449.0</td>\n",
       "      <td>78.0</td>\n",
       "      <td>3151.0</td>\n",
       "      <td>4635.0</td>\n",
       "      <td>1896.0</td>\n",
       "      <td>71.0</td>\n",
       "      <td>367.0</td>\n",
       "      <td>258.0</td>\n",
       "      <td>861.0</td>\n",
       "      <td>1480.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>Q96MU7</th>\n",
       "      <td>2981.0</td>\n",
       "      <td>8248.0</td>\n",
       "      <td>879.0</td>\n",
       "      <td>445.0</td>\n",
       "      <td>74.0</td>\n",
       "      <td>3149.0</td>\n",
       "      <td>4630.0</td>\n",
       "      <td>1892.0</td>\n",
       "      <td>68.0</td>\n",
       "      <td>363.0</td>\n",
       "      <td>254.0</td>\n",
       "      <td>859.0</td>\n",
       "      <td>1477.0</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "         S-glutathionylation  S-oxidation(nec.)  S-succination  \\\n",
       "Protein                                                          \n",
       "M0R2D3                2984.0             8251.0          882.0   \n",
       "O54824                2983.0             8248.0          882.0   \n",
       "B4E2Q0                2981.0             8248.0          879.0   \n",
       "Q60649                2983.0             8248.0          882.0   \n",
       "Q96MU7                2981.0             8248.0          879.0   \n",
       "\n",
       "         S-carbonylation  S-prenylation  S-palmitoylation  S-nitrosylation  \\\n",
       "Protein                                                                      \n",
       "M0R2D3             448.0           77.0            3151.0           4634.0   \n",
       "O54824             449.0           78.0            3153.0           4635.0   \n",
       "B4E2Q0             445.0           74.0            3148.0           4631.0   \n",
       "Q60649             449.0           78.0            3151.0           4635.0   \n",
       "Q96MU7             445.0           74.0            3149.0           4630.0   \n",
       "\n",
       "         S-sulfenylation  S-sulfonylation  S-sulfinylation  S-disulfidation  \\\n",
       "Protein                                                                       \n",
       "M0R2D3            1893.0             71.0            366.0            257.0   \n",
       "O54824            1896.0             71.0            367.0            258.0   \n",
       "B4E2Q0            1892.0             68.0            363.0            254.0   \n",
       "Q60649            1896.0             71.0            367.0            258.0   \n",
       "Q96MU7            1892.0             68.0            363.0            254.0   \n",
       "\n",
       "         S-itaconation  S-persulfidation  \n",
       "Protein                                   \n",
       "M0R2D3           862.0            1480.0  \n",
       "O54824           859.0            1480.0  \n",
       "B4E2Q0           859.0            1477.0  \n",
       "Q60649           861.0            1480.0  \n",
       "Q96MU7           859.0            1477.0  "
      ]
     },
     "execution_count": 33,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "mapping = {}\n",
    "for p,ptm,score in zip(df_res['Protein'],df_res['PTM'],df_res['Score']):\n",
    "    mapping.update({p+\":\"+ptm:score})\n",
    "\n",
    "\n",
    "Protein = list(set(df_res['Protein']))\n",
    "PTM = list(set(df_res['PTM']))\n",
    "DATA = []\n",
    "for p in Protein:\n",
    "    data =  {}\n",
    "    data.update({\"Protein\": p})\n",
    "    for ptm in PTM:\n",
    "        data.update({ptm:mapping[p+\":\"+ptm]})\n",
    "    DATA.append(data)\n",
    "    \n",
    "df = pd.DataFrame(DATA)\n",
    "df = df.set_index(\"Protein\")\n",
    "df.head()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 34,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "<AxesSubplot:ylabel='Protein'>"
      ]
     },
     "execution_count": 34,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 1000x1000 with 2 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "import matplotlib.pyplot as plt\n",
    "df = df/df.max()\n",
    "plt.figure(figsize = [10,10])\n",
    "sns.heatmap(df[0:10],cmap=\"YlGnBu\", linewidths=.5)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 35,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "(15455, 13)"
      ]
     },
     "execution_count": 35,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "df.shape"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 38,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "<AxesSubplot:>"
      ]
     },
     "execution_count": 38,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 640x480 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "df.loc[\"P12235\",:].plot.bar()"
   ]
  }
 ],
 "metadata": {
  "kernelspec": {
   "display_name": "Python 3 (ipykernel)",
   "language": "python",
   "name": "python3"
  },
  "language_info": {
   "codemirror_mode": {
    "name": "ipython",
    "version": 3
   },
   "file_extension": ".py",
   "mimetype": "text/x-python",
   "name": "python",
   "nbconvert_exporter": "python",
   "pygments_lexer": "ipython3",
   "version": "3.9.13"
  },
  "vscode": {
   "interpreter": {
    "hash": "a46d3808ce7f3f92d20203132b60af20d084112e7b3027b56a392bdfd2bca4cb"
   }
  }
 },
 "nbformat": 4,
 "nbformat_minor": 2
}