examples/l5pc/L5PC_arbor.ipynb

{
 "cells": [
  {
   "attachments": {},
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# Optimisation of a Neocortical Layer 5 Pyramidal Cell in Arbor"
   ]
  },
  {
   "attachments": {},
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "This notebook shows you how to optimise the maximal conductance of Neocortical Layer 5 Pyramidal Cell as used in Markram et al. 2015 using Arbor as the simulator.\n",
    "\n",
    "Author of this script: Werner Van Geit @ Blue Brain Project\n",
    "\n",
    "Choice of parameters, protocols and other settings was done by Etay Hay @ HUJI\n",
    "\n",
    "What's described here is a more advanced use of BluePyOpt. We suggest to first go through the introductary example here: /~https://github.com/BlueBrain/BluePyOpt/blob/master/examples/simplecell/simplecell.ipynb\n",
    "\n",
    "**If you use the methods in this notebook, we ask you to cite the following publications when publishing your research:**\n",
    "\n",
    "Van Geit, W., M. Gevaert, G. Chindemi, C. Rössert, J.-D. Courcol, E. Muller, F. Schürmann, I. Segev, and H. Markram (2016, March). BluePyOpt: Leveraging open source software and cloud infrastructure to optimise model parameters in neuroscience. ArXiv e-prints.\n",
    "http://arxiv.org/abs/1603.00500\n",
    "\n",
    "Markram, H., E. Muller, S. Ramaswamy, M. W. Reimann, M. Abdellah, C. A. Sanchez, A. Ailamaki, L. Alonso-Nanclares, N. Antille, S. Arsever, et al. (2015). Reconstruction and simulation of neocortical microcircuitry. Cell 163(2), 456–492.\n",
    "http://www.cell.com/abstract/S0092-8674%2815%2901191-5\n",
    "\n",
    "Some of the modules loaded in this script are located in the L5PC example folder: /~https://github.com/BlueBrain/BluePyOpt/tree/master/examples/l5pc "
   ]
  },
  {
   "attachments": {},
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "We first load the bluepyopt python module, the ephys submodule and some helper functionality"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 1,
   "metadata": {
    "collapsed": false,
    "jupyter": {
     "outputs_hidden": false
    }
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "/home/lukasd/src/arbor/dev/BluePyOpt/examples/l5pc\n",
      "Mod files: \"mechanisms/CaDynamics_E2.mod\" \"mechanisms/Ca_HVA.mod\" \"mechanisms/Ca_LVAst.mod\" \"mechanisms/Ih.mod\" \"mechanisms/Im.mod\" \"mechanisms/K_Pst.mod\" \"mechanisms/K_Tst.mod\" \"mechanisms/Nap_Et2.mod\" \"mechanisms/NaTa_t.mod\" \"mechanisms/NaTs2_t.mod\" \"mechanisms/SK_E2.mod\" \"mechanisms/SKv3_1.mod\"\n",
      "\n",
      "Creating x86_64 directory for .o files.\n",
      "\n",
      "COBJS=''\n",
      " -> \u001b[32mCompiling\u001b[0m mod_func.c\n",
      " -> \u001b[32mNMODL\u001b[0m ../mechanisms/CaDynamics_E2.mod\n",
      "x86_64-linux-gnu-gcc -O2   -I.   -I/home/lukasd/src/arbor/dev/venv/lib/python3.8/site-packages/neuron/.data/include  -I/nrnwheel/openmpi/include -fPIC -c mod_func.c -o mod_func.o\n",
      "(cd \"../mechanisms\"; MODLUNIT=/home/lukasd/src/arbor/dev/venv/lib/python3.8/site-packages/neuron/.data/share/nrn/lib/nrnunits.lib /home/lukasd/src/arbor/dev/venv/lib/python3.8/site-packages/neuron/.data/bin/nocmodl CaDynamics_E2.mod -o \"/home/lukasd/src/arbor/dev/BluePyOpt/examples/l5pc/x86_64\")\n",
      " -> \u001b[32mNMODL\u001b[0m ../mechanisms/Ca_HVA.mod\n",
      "(cd \"../mechanisms\"; MODLUNIT=/home/lukasd/src/arbor/dev/venv/lib/python3.8/site-packages/neuron/.data/share/nrn/lib/nrnunits.lib /home/lukasd/src/arbor/dev/venv/lib/python3.8/site-packages/neuron/.data/bin/nocmodl Ca_HVA.mod -o \"/home/lukasd/src/arbor/dev/BluePyOpt/examples/l5pc/x86_64\")\n",
      " -> \u001b[32mNMODL\u001b[0m ../mechanisms/Ca_LVAst.mod\n",
      "(cd \"../mechanisms\"; MODLUNIT=/home/lukasd/src/arbor/dev/venv/lib/python3.8/site-packages/neuron/.data/share/nrn/lib/nrnunits.lib /home/lukasd/src/arbor/dev/venv/lib/python3.8/site-packages/neuron/.data/bin/nocmodl Ca_LVAst.mod -o \"/home/lukasd/src/arbor/dev/BluePyOpt/examples/l5pc/x86_64\")\n",
      "Translating CaDynamics_E2.mod into /home/lukasd/src/arbor/dev/BluePyOpt/examples/l5pc/x86_64/CaDynamics_E2.c\n",
      "Translating Ca_HVA.mod into /home/lukasd/src/arbor/dev/BluePyOpt/examples/l5pc/x86_64/Ca_HVA.c\n",
      "Translating Ca_LVAst.mod into /home/lukasd/src/arbor/dev/BluePyOpt/examples/l5pc/x86_64/Ca_LVAst.c\n",
      "Thread Safe\n",
      "Thread Safe\n",
      "Thread Safe\n",
      " -> \u001b[32mNMODL\u001b[0m ../mechanisms/Ih.mod\n",
      "(cd \"../mechanisms\"; MODLUNIT=/home/lukasd/src/arbor/dev/venv/lib/python3.8/site-packages/neuron/.data/share/nrn/lib/nrnunits.lib /home/lukasd/src/arbor/dev/venv/lib/python3.8/site-packages/neuron/.data/bin/nocmodl Ih.mod -o \"/home/lukasd/src/arbor/dev/BluePyOpt/examples/l5pc/x86_64\")\n",
      "Translating Ih.mod into /home/lukasd/src/arbor/dev/BluePyOpt/examples/l5pc/x86_64/Ih.c\n",
      "Thread Safe\n",
      " -> \u001b[32mNMODL\u001b[0m ../mechanisms/Im.mod\n",
      "(cd \"../mechanisms\"; MODLUNIT=/home/lukasd/src/arbor/dev/venv/lib/python3.8/site-packages/neuron/.data/share/nrn/lib/nrnunits.lib /home/lukasd/src/arbor/dev/venv/lib/python3.8/site-packages/neuron/.data/bin/nocmodl Im.mod -o \"/home/lukasd/src/arbor/dev/BluePyOpt/examples/l5pc/x86_64\")\n",
      "Translating Im.mod into /home/lukasd/src/arbor/dev/BluePyOpt/examples/l5pc/x86_64/Im.c\n",
      "Thread Safe\n",
      " -> \u001b[32mNMODL\u001b[0m ../mechanisms/K_Pst.mod\n",
      "(cd \"../mechanisms\"; MODLUNIT=/home/lukasd/src/arbor/dev/venv/lib/python3.8/site-packages/neuron/.data/share/nrn/lib/nrnunits.lib /home/lukasd/src/arbor/dev/venv/lib/python3.8/site-packages/neuron/.data/bin/nocmodl K_Pst.mod -o \"/home/lukasd/src/arbor/dev/BluePyOpt/examples/l5pc/x86_64\")\n",
      " -> \u001b[32mNMODL\u001b[0m ../mechanisms/K_Tst.mod\n",
      "(cd \"../mechanisms\"; MODLUNIT=/home/lukasd/src/arbor/dev/venv/lib/python3.8/site-packages/neuron/.data/share/nrn/lib/nrnunits.lib /home/lukasd/src/arbor/dev/venv/lib/python3.8/site-packages/neuron/.data/bin/nocmodl K_Tst.mod -o \"/home/lukasd/src/arbor/dev/BluePyOpt/examples/l5pc/x86_64\")\n",
      " -> \u001b[32mNMODL\u001b[0m ../mechanisms/Nap_Et2.mod\n",
      "Translating K_Pst.mod into /home/lukasd/src/arbor/dev/BluePyOpt/examples/l5pc/x86_64/K_Pst.c\n",
      "Thread Safe\n",
      "(cd \"../mechanisms\"; MODLUNIT=/home/lukasd/src/arbor/dev/venv/lib/python3.8/site-packages/neuron/.data/share/nrn/lib/nrnunits.lib /home/lukasd/src/arbor/dev/venv/lib/python3.8/site-packages/neuron/.data/bin/nocmodl Nap_Et2.mod -o \"/home/lukasd/src/arbor/dev/BluePyOpt/examples/l5pc/x86_64\")\n",
      "Translating K_Tst.mod into /home/lukasd/src/arbor/dev/BluePyOpt/examples/l5pc/x86_64/K_Tst.c\n",
      "Thread Safe\n",
      "Translating Nap_Et2.mod into /home/lukasd/src/arbor/dev/BluePyOpt/examples/l5pc/x86_64/Nap_Et2.c\n",
      "Thread Safe\n",
      " -> \u001b[32mNMODL\u001b[0m ../mechanisms/NaTa_t.mod\n",
      "(cd \"../mechanisms\"; MODLUNIT=/home/lukasd/src/arbor/dev/venv/lib/python3.8/site-packages/neuron/.data/share/nrn/lib/nrnunits.lib /home/lukasd/src/arbor/dev/venv/lib/python3.8/site-packages/neuron/.data/bin/nocmodl NaTa_t.mod -o \"/home/lukasd/src/arbor/dev/BluePyOpt/examples/l5pc/x86_64\")\n",
      " -> \u001b[32mNMODL\u001b[0m ../mechanisms/NaTs2_t.mod\n",
      " -> \u001b[32mNMODL\u001b[0m ../mechanisms/SK_E2.mod\n",
      "(cd \"../mechanisms\"; MODLUNIT=/home/lukasd/src/arbor/dev/venv/lib/python3.8/site-packages/neuron/.data/share/nrn/lib/nrnunits.lib /home/lukasd/src/arbor/dev/venv/lib/python3.8/site-packages/neuron/.data/bin/nocmodl NaTs2_t.mod -o \"/home/lukasd/src/arbor/dev/BluePyOpt/examples/l5pc/x86_64\")\n",
      "(cd \"../mechanisms\"; MODLUNIT=/home/lukasd/src/arbor/dev/venv/lib/python3.8/site-packages/neuron/.data/share/nrn/lib/nrnunits.lib /home/lukasd/src/arbor/dev/venv/lib/python3.8/site-packages/neuron/.data/bin/nocmodl SK_E2.mod -o \"/home/lukasd/src/arbor/dev/BluePyOpt/examples/l5pc/x86_64\")\n",
      "Translating NaTa_t.mod into /home/lukasd/src/arbor/dev/BluePyOpt/examples/l5pc/x86_64/NaTa_t.c\n",
      "Translating NaTs2_t.mod into /home/lukasd/src/arbor/dev/BluePyOpt/examples/l5pc/x86_64/NaTs2_t.c\n",
      "Thread Safe\n",
      "Thread Safe\n",
      " -> \u001b[32mNMODL\u001b[0m ../mechanisms/SKv3_1.mod\n",
      "(cd \"../mechanisms\"; MODLUNIT=/home/lukasd/src/arbor/dev/venv/lib/python3.8/site-packages/neuron/.data/share/nrn/lib/nrnunits.lib /home/lukasd/src/arbor/dev/venv/lib/python3.8/site-packages/neuron/.data/bin/nocmodl SKv3_1.mod -o \"/home/lukasd/src/arbor/dev/BluePyOpt/examples/l5pc/x86_64\")\n",
      "Translating SK_E2.mod into /home/lukasd/src/arbor/dev/BluePyOpt/examples/l5pc/x86_64/SK_E2.c\n",
      "Thread Safe\n",
      "Translating SKv3_1.mod into /home/lukasd/src/arbor/dev/BluePyOpt/examples/l5pc/x86_64/SKv3_1.c\n",
      " -> \u001b[32mCompiling\u001b[0m CaDynamics_E2.c\n",
      "x86_64-linux-gnu-gcc -O2   -I\"../mechanisms\" -I.   -I/home/lukasd/src/arbor/dev/venv/lib/python3.8/site-packages/neuron/.data/include  -I/nrnwheel/openmpi/include -fPIC -c CaDynamics_E2.c -o CaDynamics_E2.o\n",
      " -> \u001b[32mCompiling\u001b[0m Ca_HVA.c\n",
      "x86_64-linux-gnu-gcc -O2   -I\"../mechanisms\" -I.   -I/home/lukasd/src/arbor/dev/venv/lib/python3.8/site-packages/neuron/.data/include  -I/nrnwheel/openmpi/include -fPIC -c Ca_HVA.c -o Ca_HVA.o\n",
      "Thread Safe\n",
      " -> \u001b[32mCompiling\u001b[0m Ca_LVAst.c\n",
      "x86_64-linux-gnu-gcc -O2   -I\"../mechanisms\" -I.   -I/home/lukasd/src/arbor/dev/venv/lib/python3.8/site-packages/neuron/.data/include  -I/nrnwheel/openmpi/include -fPIC -c Ca_LVAst.c -o Ca_LVAst.o\n",
      " -> \u001b[32mCompiling\u001b[0m Ih.c\n",
      "x86_64-linux-gnu-gcc -O2   -I\"../mechanisms\" -I.   -I/home/lukasd/src/arbor/dev/venv/lib/python3.8/site-packages/neuron/.data/include  -I/nrnwheel/openmpi/include -fPIC -c Ih.c -o Ih.o\n",
      " -> \u001b[32mCompiling\u001b[0m Im.c\n",
      "x86_64-linux-gnu-gcc -O2   -I\"../mechanisms\" -I.   -I/home/lukasd/src/arbor/dev/venv/lib/python3.8/site-packages/neuron/.data/include  -I/nrnwheel/openmpi/include -fPIC -c Im.c -o Im.o\n",
      " -> \u001b[32mCompiling\u001b[0m K_Pst.c\n",
      "x86_64-linux-gnu-gcc -O2   -I\"../mechanisms\" -I.   -I/home/lukasd/src/arbor/dev/venv/lib/python3.8/site-packages/neuron/.data/include  -I/nrnwheel/openmpi/include -fPIC -c K_Pst.c -o K_Pst.o\n",
      " -> \u001b[32mCompiling\u001b[0m K_Tst.c\n",
      "x86_64-linux-gnu-gcc -O2   -I\"../mechanisms\" -I.   -I/home/lukasd/src/arbor/dev/venv/lib/python3.8/site-packages/neuron/.data/include  -I/nrnwheel/openmpi/include -fPIC -c K_Tst.c -o K_Tst.o\n",
      " -> \u001b[32mCompiling\u001b[0m Nap_Et2.c\n",
      "x86_64-linux-gnu-gcc -O2   -I\"../mechanisms\" -I.   -I/home/lukasd/src/arbor/dev/venv/lib/python3.8/site-packages/neuron/.data/include  -I/nrnwheel/openmpi/include -fPIC -c Nap_Et2.c -o Nap_Et2.o\n",
      " -> \u001b[32mCompiling\u001b[0m NaTa_t.c\n",
      "x86_64-linux-gnu-gcc -O2   -I\"../mechanisms\" -I.   -I/home/lukasd/src/arbor/dev/venv/lib/python3.8/site-packages/neuron/.data/include  -I/nrnwheel/openmpi/include -fPIC -c NaTa_t.c -o NaTa_t.o\n",
      " -> \u001b[32mCompiling\u001b[0m NaTs2_t.c\n",
      "x86_64-linux-gnu-gcc -O2   -I\"../mechanisms\" -I.   -I/home/lukasd/src/arbor/dev/venv/lib/python3.8/site-packages/neuron/.data/include  -I/nrnwheel/openmpi/include -fPIC -c NaTs2_t.c -o NaTs2_t.o\n",
      " -> \u001b[32mCompiling\u001b[0m SK_E2.c\n",
      "x86_64-linux-gnu-gcc -O2   -I\"../mechanisms\" -I.   -I/home/lukasd/src/arbor/dev/venv/lib/python3.8/site-packages/neuron/.data/include  -I/nrnwheel/openmpi/include -fPIC -c SK_E2.c -o SK_E2.o\n",
      " -> \u001b[32mCompiling\u001b[0m SKv3_1.c\n",
      "x86_64-linux-gnu-gcc -O2   -I\"../mechanisms\" -I.   -I/home/lukasd/src/arbor/dev/venv/lib/python3.8/site-packages/neuron/.data/include  -I/nrnwheel/openmpi/include -fPIC -c SKv3_1.c -o SKv3_1.o\n",
      " => \u001b[32mLINKING\u001b[0m shared library ./libnrnmech.so\n",
      "x86_64-linux-gnu-g++ -O2 -DVERSION_INFO='8.0.2' -std=c++11 -shared -fPIC  -I /home/lukasd/src/arbor/dev/venv/lib/python3.8/site-packages/neuron/.data/include -o ./libnrnmech.so -Wl,-soname,libnrnmech.so \\\n",
      "  ./mod_func.o ./CaDynamics_E2.o ./Ca_HVA.o ./Ca_LVAst.o ./Ih.o ./Im.o ./K_Pst.o ./K_Tst.o ./Nap_Et2.o ./NaTa_t.o ./NaTs2_t.o ./SK_E2.o ./SKv3_1.o  -L/home/lukasd/src/arbor/dev/venv/lib/python3.8/site-packages/neuron/.data/lib -lnrniv -Wl,-rpath,/home/lukasd/src/arbor/dev/venv/lib/python3.8/site-packages/neuron/.data/lib   \n",
      "rm -f ./.libs/libnrnmech.so ; mkdir -p ./.libs ; cp ./libnrnmech.so ./.libs/libnrnmech.so\n",
      "Successfully created x86_64/special\n"
     ]
    }
   ],
   "source": [
    "%load_ext autoreload\n",
    "%autoreload\n",
    "\n",
    "!nrnivmodl mechanisms\n",
    "import bluepyopt as bpopt\n",
    "import bluepyopt.ephys as ephys\n",
    "\n",
    "import pprint\n",
    "pp = pprint.PrettyPrinter(indent=2)\n",
    "\n",
    "%matplotlib inline\n",
    "import matplotlib.pyplot as plt"
   ]
  },
  {
   "attachments": {},
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Enable the code below to enable debug level logging"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "metadata": {},
   "outputs": [],
   "source": [
    "# import logging                                                                      \n",
    "# logging.basicConfig()                                                               \n",
    "# logger = logging.getLogger()                                                        \n",
    "# logger.setLevel(logging.DEBUG)   "
   ]
  },
  {
   "attachments": {},
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Model description"
   ]
  },
  {
   "attachments": {},
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Morphology"
   ]
  },
  {
   "attachments": {},
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "We're using a complex reconstructed morphology of an L5PC cell. Let's visualise this with the BlueBrain NeuroM software. Alternatively, the cell model can be exported to JSON/ACC by running\n",
    "\n",
    "```shell\n",
    "./generate_acc.py --replace-axon --output <output-dir>\n",
    "```\n",
    "\n",
    "The output can be visualized with the Arbor GUI (graphical user interface) as shown in the [documentation](https://docs.arbor-sim.org/en/latest/tutorial/single_cell_bluepyopt.html)."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "metadata": {
    "collapsed": false,
    "jupyter": {
     "outputs_hidden": false
    }
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Requirement already satisfied: neurom in /home/lukasd/src/arbor/dev/venv/lib/python3.8/site-packages (3.2.2)\n",
      "Requirement already satisfied: morphio>=3.1.1 in /home/lukasd/src/arbor/dev/venv/lib/python3.8/site-packages (from neurom) (3.3.3)\n",
      "Requirement already satisfied: numpy>=1.8.0 in /home/lukasd/src/arbor/dev/venv/lib/python3.8/site-packages (from neurom) (1.22.3)\n",
      "Requirement already satisfied: scipy>=1.2.0 in /home/lukasd/src/arbor/dev/venv/lib/python3.8/site-packages (from neurom) (1.8.0)\n",
      "Requirement already satisfied: matplotlib>=3.2.1 in /home/lukasd/src/arbor/dev/venv/lib/python3.8/site-packages (from neurom) (3.5.1)\n",
      "Requirement already satisfied: pandas>=1.0.5 in /home/lukasd/src/arbor/dev/venv/lib/python3.8/site-packages (from neurom) (1.4.1)\n",
      "Requirement already satisfied: tqdm>=4.8.4 in /home/lukasd/src/arbor/dev/venv/lib/python3.8/site-packages (from neurom) (4.63.1)\n",
      "Requirement already satisfied: click>=7.0 in /home/lukasd/src/arbor/dev/venv/lib/python3.8/site-packages (from neurom) (8.1.3)\n",
      "Requirement already satisfied: pyyaml>=3.10 in /home/lukasd/src/arbor/dev/venv/lib/python3.8/site-packages (from neurom) (6.0)\n",
      "Requirement already satisfied: pillow>=6.2.0 in /home/lukasd/src/arbor/dev/venv/lib/python3.8/site-packages (from matplotlib>=3.2.1->neurom) (9.0.1)\n",
      "Requirement already satisfied: packaging>=20.0 in /home/lukasd/src/arbor/dev/venv/lib/python3.8/site-packages (from matplotlib>=3.2.1->neurom) (21.3)\n",
      "Requirement already satisfied: pyparsing>=2.2.1 in /home/lukasd/src/arbor/dev/venv/lib/python3.8/site-packages (from matplotlib>=3.2.1->neurom) (3.0.7)\n",
      "Requirement already satisfied: python-dateutil>=2.7 in /home/lukasd/src/arbor/dev/venv/lib/python3.8/site-packages (from matplotlib>=3.2.1->neurom) (2.8.2)\n",
      "Requirement already satisfied: fonttools>=4.22.0 in /home/lukasd/src/arbor/dev/venv/lib/python3.8/site-packages (from matplotlib>=3.2.1->neurom) (4.31.2)\n",
      "Requirement already satisfied: cycler>=0.10 in /home/lukasd/src/arbor/dev/venv/lib/python3.8/site-packages (from matplotlib>=3.2.1->neurom) (0.11.0)\n",
      "Requirement already satisfied: kiwisolver>=1.0.1 in /home/lukasd/src/arbor/dev/venv/lib/python3.8/site-packages (from matplotlib>=3.2.1->neurom) (1.4.2)\n",
      "Requirement already satisfied: pytz>=2020.1 in /home/lukasd/src/arbor/dev/venv/lib/python3.8/site-packages (from pandas>=1.0.5->neurom) (2022.1)\n",
      "Requirement already satisfied: six>=1.5 in /home/lukasd/src/arbor/dev/venv/lib/python3.8/site-packages (from python-dateutil>=2.7->matplotlib>=3.2.1->neurom) (1.16.0)\n",
      "\n",
      "\u001b[1m[\u001b[0m\u001b[34;49mnotice\u001b[0m\u001b[1;39;49m]\u001b[0m\u001b[39;49m A new release of pip available: \u001b[0m\u001b[31;49m22.2.2\u001b[0m\u001b[39;49m -> \u001b[0m\u001b[32;49m22.3\u001b[0m\n",
      "\u001b[1m[\u001b[0m\u001b[34;49mnotice\u001b[0m\u001b[1;39;49m]\u001b[0m\u001b[39;49m To update, run: \u001b[0m\u001b[32;49mpip install --upgrade pip\u001b[0m\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "/tmp/ipykernel_466037/1031438697.py:3: NeuroMDeprecationWarning: `neurom.io.utils.load_neuron` is deprecated in favor of `neurom.io.utils.load_morphology`\n",
      "  neurom.viewer.draw(neurom.load_neuron('morphology/C060114A7.asc'));\n"
     ]
    },
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "!pip install neurom --upgrade\n",
    "import neurom.viewer\n",
    "neurom.viewer.draw(neurom.load_neuron('morphology/C060114A7.asc'));"
   ]
  },
  {
   "attachments": {},
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "To load the morphology we create a NrnFileMorphology object. We set 'do_replace_axon' to True to replace the axon with a AIS."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "metadata": {
    "collapsed": false,
    "jupyter": {
     "outputs_hidden": false
    }
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "morphology/C060114A7.asc\n"
     ]
    }
   ],
   "source": [
    "morphology = ephys.morphologies.NrnFileMorphology('morphology/C060114A7.asc', do_replace_axon=True)\n",
    "print(str(morphology))"
   ]
  },
  {
   "attachments": {},
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Parameters"
   ]
  },
  {
   "attachments": {},
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Since we have many parameters in this model, they are stored in a json file: /~https://github.com/BlueBrain/BluePyOpt/blob/master/examples/l5pc/config/parameters.json"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "metadata": {
    "collapsed": false,
    "jupyter": {
     "outputs_hidden": false
    }
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "['g_pas', 'e_pas', 'cm', 'Ra', 'v_init', 'celsius', 'ena', 'ek', 'cm', 'ena', 'ek', 'cm', 'ena', 'ek', 'gIhbar_Ih', 'gNaTs2_tbar_NaTs2_t', 'gSKv3_1bar_SKv3_1', 'gImbar_Im', 'gIhbar_Ih', 'gNaTa_tbar_NaTa_t', 'gNap_Et2bar_Nap_Et2', 'gK_Pstbar_K_Pst', 'gK_Tstbar_K_Tst', 'gSK_E2bar_SK_E2', 'gSKv3_1bar_SKv3_1', 'gCa_HVAbar_Ca_HVA', 'gCa_LVAstbar_Ca_LVAst', 'gamma_CaDynamics_E2', 'decay_CaDynamics_E2', 'gNaTs2_tbar_NaTs2_t', 'gSKv3_1bar_SKv3_1', 'gSK_E2bar_SK_E2', 'gCa_HVAbar_Ca_HVA', 'gCa_LVAstbar_Ca_LVAst', 'gamma_CaDynamics_E2', 'decay_CaDynamics_E2', 'gIhbar_Ih']\n"
     ]
    }
   ],
   "source": [
    "import json\n",
    "param_configs = json.load(open('config/parameters.json'))\n",
    "print([param_config['param_name'] for param_config in param_configs])"
   ]
  },
  {
   "attachments": {},
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "The directory that contains this notebook has a module that will load all the parameters in BluePyOpt Parameter objects"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 6,
   "metadata": {
    "collapsed": false,
    "jupyter": {
     "outputs_hidden": false
    }
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "g_pas.all: ['all'] g_pas = 3e-05\n",
      "e_pas.all: ['all'] e_pas = -75\n",
      "cm.all: ['all'] cm = 1\n",
      "Ra.all: ['all'] Ra = 100\n",
      "v_init: v_init = -65\n",
      "celsius: celsius = 34\n",
      "ena.apical: ['apical'] ena = 50\n",
      "ek.apical: ['apical'] ek = -85\n",
      "cm.apical: ['apical'] cm = 2\n",
      "ena.somatic: ['somatic'] ena = 50\n",
      "ek.somatic: ['somatic'] ek = -85\n",
      "cm.basal: ['basal'] cm = 2\n",
      "ena.axonal: ['axonal'] ena = 50\n",
      "ek.axonal: ['axonal'] ek = -85\n",
      "gIhbar_Ih.basal: ['basal'] gIhbar_Ih = 8e-05\n",
      "gNaTs2_tbar_NaTs2_t.apical: ['apical'] gNaTs2_tbar_NaTs2_t = [0, 0.04]\n",
      "gSKv3_1bar_SKv3_1.apical: ['apical'] gSKv3_1bar_SKv3_1 = [0, 0.04]\n",
      "gImbar_Im.apical: ['apical'] gImbar_Im = [0, 0.001]\n",
      "gIhbar_Ih.apical: ['apical'] gIhbar_Ih = 8e-05\n",
      "gNaTa_tbar_NaTa_t.axonal: ['axonal'] gNaTa_tbar_NaTa_t = [0, 4]\n",
      "gNap_Et2bar_Nap_Et2.axonal: ['axonal'] gNap_Et2bar_Nap_Et2 = [0, 4]\n",
      "gK_Pstbar_K_Pst.axonal: ['axonal'] gK_Pstbar_K_Pst = [0, 1]\n",
      "gK_Tstbar_K_Tst.axonal: ['axonal'] gK_Tstbar_K_Tst = [0, 0.1]\n",
      "gSK_E2bar_SK_E2.axonal: ['axonal'] gSK_E2bar_SK_E2 = [0, 0.1]\n",
      "gSKv3_1bar_SKv3_1.axonal: ['axonal'] gSKv3_1bar_SKv3_1 = [0, 2]\n",
      "gCa_HVAbar_Ca_HVA.axonal: ['axonal'] gCa_HVAbar_Ca_HVA = [0, 0.001]\n",
      "gCa_LVAstbar_Ca_LVAst.axonal: ['axonal'] gCa_LVAstbar_Ca_LVAst = [0, 0.01]\n",
      "gamma_CaDynamics_E2.axonal: ['axonal'] gamma_CaDynamics_E2 = [0.0005, 0.05]\n",
      "decay_CaDynamics_E2.axonal: ['axonal'] decay_CaDynamics_E2 = [20, 1000]\n",
      "gNaTs2_tbar_NaTs2_t.somatic: ['somatic'] gNaTs2_tbar_NaTs2_t = [0, 1]\n",
      "gSKv3_1bar_SKv3_1.somatic: ['somatic'] gSKv3_1bar_SKv3_1 = [0, 1]\n",
      "gSK_E2bar_SK_E2.somatic: ['somatic'] gSK_E2bar_SK_E2 = [0, 0.1]\n",
      "gCa_HVAbar_Ca_HVA.somatic: ['somatic'] gCa_HVAbar_Ca_HVA = [0, 0.001]\n",
      "gCa_LVAstbar_Ca_LVAst.somatic: ['somatic'] gCa_LVAstbar_Ca_LVAst = [0, 0.01]\n",
      "gamma_CaDynamics_E2.somatic: ['somatic'] gamma_CaDynamics_E2 = [0.0005, 0.05]\n",
      "decay_CaDynamics_E2.somatic: ['somatic'] decay_CaDynamics_E2 = [20, 1000]\n",
      "gIhbar_Ih.somatic: ['somatic'] gIhbar_Ih = 8e-05\n"
     ]
    }
   ],
   "source": [
    "import l5pc_model\n",
    "parameters = l5pc_model.define_parameters()\n",
    "print('\\n'.join('%s' % param for param in parameters))"
   ]
  },
  {
   "attachments": {},
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "As you can see there are two types of parameters, parameters with a fixed value and parameters with bounds. The latter will be optimised by the algorithm."
   ]
  },
  {
   "attachments": {},
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Mechanism"
   ]
  },
  {
   "attachments": {},
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "We also need to add all the necessary mechanisms, like ion channels to the model. \n",
    "The configuration of the mechanisms is also stored in a json file, and can be loaded in a similar way."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 7,
   "metadata": {
    "collapsed": false,
    "jupyter": {
     "outputs_hidden": false
    }
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Ih.basal: Ih at ['basal']\n",
      "pas.all: pas at ['all']\n",
      "Ih.apical: Ih at ['apical']\n",
      "Im.apical: Im at ['apical']\n",
      "SKv3_1.apical: SKv3_1 at ['apical']\n",
      "NaTs2_t.apical: NaTs2_t at ['apical']\n",
      "Ca_LVAst.axonal: Ca_LVAst at ['axonal']\n",
      "Ca_HVA.axonal: Ca_HVA at ['axonal']\n",
      "CaDynamics_E2.axonal: CaDynamics_E2 at ['axonal']\n",
      "SKv3_1.axonal: SKv3_1 at ['axonal']\n",
      "SK_E2.axonal: SK_E2 at ['axonal']\n",
      "K_Tst.axonal: K_Tst at ['axonal']\n",
      "K_Pst.axonal: K_Pst at ['axonal']\n",
      "Nap_Et2.axonal: Nap_Et2 at ['axonal']\n",
      "NaTa_t.axonal: NaTa_t at ['axonal']\n",
      "NaTs2_t.somatic: NaTs2_t at ['somatic']\n",
      "SKv3_1.somatic: SKv3_1 at ['somatic']\n",
      "SK_E2.somatic: SK_E2 at ['somatic']\n",
      "CaDynamics_E2.somatic: CaDynamics_E2 at ['somatic']\n",
      "Ca_HVA.somatic: Ca_HVA at ['somatic']\n",
      "Ca_LVAst.somatic: Ca_LVAst at ['somatic']\n",
      "Ih.somatic: Ih at ['somatic']\n"
     ]
    }
   ],
   "source": [
    "mechanisms = l5pc_model.define_mechanisms()\n",
    "print('\\n'.join('%s' % mech for mech in mechanisms))"
   ]
  },
  {
   "attachments": {},
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# Cell model"
   ]
  },
  {
   "attachments": {},
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "With the morphology, mechanisms and parameters we can build the cell model. If we use axon-replacement, we must 3d-instantiate the morphology first in the Neuron simulator to obtain a faithful representation in Arbor.\n",
    "\n",
    "Note that before `l5pc_cell` can subsequently be used in a Neuron protocol, `l5pc_cell.destroy(sim=nrn_sim)` must be invoked. "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 8,
   "metadata": {
    "collapsed": false,
    "jupyter": {
     "outputs_hidden": false
    }
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "l5pc:\n",
      "  morphology:\n",
      "    morphology/C060114A7.asc\n",
      "  mechanisms:\n",
      "    Ih.basal: Ih at ['basal']\n",
      "    pas.all: pas at ['all']\n",
      "    Ih.apical: Ih at ['apical']\n",
      "    Im.apical: Im at ['apical']\n",
      "    SKv3_1.apical: SKv3_1 at ['apical']\n",
      "    NaTs2_t.apical: NaTs2_t at ['apical']\n",
      "    Ca_LVAst.axonal: Ca_LVAst at ['axonal']\n",
      "    Ca_HVA.axonal: Ca_HVA at ['axonal']\n",
      "    CaDynamics_E2.axonal: CaDynamics_E2 at ['axonal']\n",
      "    SKv3_1.axonal: SKv3_1 at ['axonal']\n",
      "    SK_E2.axonal: SK_E2 at ['axonal']\n",
      "    K_Tst.axonal: K_Tst at ['axonal']\n",
      "    K_Pst.axonal: K_Pst at ['axonal']\n",
      "    Nap_Et2.axonal: Nap_Et2 at ['axonal']\n",
      "    NaTa_t.axonal: NaTa_t at ['axonal']\n",
      "    NaTs2_t.somatic: NaTs2_t at ['somatic']\n",
      "    SKv3_1.somatic: SKv3_1 at ['somatic']\n",
      "    SK_E2.somatic: SK_E2 at ['somatic']\n",
      "    CaDynamics_E2.somatic: CaDynamics_E2 at ['somatic']\n",
      "    Ca_HVA.somatic: Ca_HVA at ['somatic']\n",
      "    Ca_LVAst.somatic: Ca_LVAst at ['somatic']\n",
      "    Ih.somatic: Ih at ['somatic']\n",
      "  params:\n",
      "    g_pas.all: ['all'] g_pas = 3e-05\n",
      "    e_pas.all: ['all'] e_pas = -75\n",
      "    cm.all: ['all'] cm = 1\n",
      "    Ra.all: ['all'] Ra = 100\n",
      "    v_init: v_init = -65\n",
      "    celsius: celsius = 34\n",
      "    ena.apical: ['apical'] ena = 50\n",
      "    ek.apical: ['apical'] ek = -85\n",
      "    cm.apical: ['apical'] cm = 2\n",
      "    ena.somatic: ['somatic'] ena = 50\n",
      "    ek.somatic: ['somatic'] ek = -85\n",
      "    cm.basal: ['basal'] cm = 2\n",
      "    ena.axonal: ['axonal'] ena = 50\n",
      "    ek.axonal: ['axonal'] ek = -85\n",
      "    gIhbar_Ih.basal: ['basal'] gIhbar_Ih = 8e-05\n",
      "    gNaTs2_tbar_NaTs2_t.apical: ['apical'] gNaTs2_tbar_NaTs2_t = [0, 0.04]\n",
      "    gSKv3_1bar_SKv3_1.apical: ['apical'] gSKv3_1bar_SKv3_1 = [0, 0.04]\n",
      "    gImbar_Im.apical: ['apical'] gImbar_Im = [0, 0.001]\n",
      "    gIhbar_Ih.apical: ['apical'] gIhbar_Ih = 8e-05\n",
      "    gNaTa_tbar_NaTa_t.axonal: ['axonal'] gNaTa_tbar_NaTa_t = [0, 4]\n",
      "    gNap_Et2bar_Nap_Et2.axonal: ['axonal'] gNap_Et2bar_Nap_Et2 = [0, 4]\n",
      "    gK_Pstbar_K_Pst.axonal: ['axonal'] gK_Pstbar_K_Pst = [0, 1]\n",
      "    gK_Tstbar_K_Tst.axonal: ['axonal'] gK_Tstbar_K_Tst = [0, 0.1]\n",
      "    gSK_E2bar_SK_E2.axonal: ['axonal'] gSK_E2bar_SK_E2 = [0, 0.1]\n",
      "    gSKv3_1bar_SKv3_1.axonal: ['axonal'] gSKv3_1bar_SKv3_1 = [0, 2]\n",
      "    gCa_HVAbar_Ca_HVA.axonal: ['axonal'] gCa_HVAbar_Ca_HVA = [0, 0.001]\n",
      "    gCa_LVAstbar_Ca_LVAst.axonal: ['axonal'] gCa_LVAstbar_Ca_LVAst = [0, 0.01]\n",
      "    gamma_CaDynamics_E2.axonal: ['axonal'] gamma_CaDynamics_E2 = [0.0005, 0.05]\n",
      "    decay_CaDynamics_E2.axonal: ['axonal'] decay_CaDynamics_E2 = [20, 1000]\n",
      "    gNaTs2_tbar_NaTs2_t.somatic: ['somatic'] gNaTs2_tbar_NaTs2_t = [0, 1]\n",
      "    gSKv3_1bar_SKv3_1.somatic: ['somatic'] gSKv3_1bar_SKv3_1 = [0, 1]\n",
      "    gSK_E2bar_SK_E2.somatic: ['somatic'] gSK_E2bar_SK_E2 = [0, 0.1]\n",
      "    gCa_HVAbar_Ca_HVA.somatic: ['somatic'] gCa_HVAbar_Ca_HVA = [0, 0.001]\n",
      "    gCa_LVAstbar_Ca_LVAst.somatic: ['somatic'] gCa_LVAstbar_Ca_LVAst = [0, 0.01]\n",
      "    gamma_CaDynamics_E2.somatic: ['somatic'] gamma_CaDynamics_E2 = [0.0005, 0.05]\n",
      "    decay_CaDynamics_E2.somatic: ['somatic'] decay_CaDynamics_E2 = [20, 1000]\n",
      "    gIhbar_Ih.somatic: ['somatic'] gIhbar_Ih = 8e-05\n",
      "\n"
     ]
    }
   ],
   "source": [
    "l5pc_cell = ephys.models.CellModel('l5pc', morph=morphology, mechs=mechanisms, params=parameters)\n",
    "\n",
    "if morphology.do_replace_axon:\n",
    "    nrn_sim = ephys.simulators.NrnSimulator()\n",
    "    # invoke this before exporting an axon-replaced morphology to JSON/ACC-format\n",
    "    l5pc_cell.instantiate_morphology_3d(nrn_sim)\n",
    "\n",
    "print(l5pc_cell)"
   ]
  },
  {
   "attachments": {},
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "For use in the cell evaluator later, we need to make a list of the name of the parameters we are going to optimise.\n",
    "These are the parameters that are not frozen."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 9,
   "metadata": {},
   "outputs": [],
   "source": [
    "param_names = [param.name for param in l5pc_cell.params.values() if not param.frozen]      "
   ]
  },
  {
   "attachments": {},
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Protocols"
   ]
  },
  {
   "attachments": {},
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Now that we have a cell model, we can apply protocols to it. The protocols are also stored in a json file."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 10,
   "metadata": {
    "collapsed": false,
    "jupyter": {
     "outputs_hidden": false
    }
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "{'bAP': {'stimuli': [{'delay': 295, 'amp': 1.9, 'duration': 5, 'totduration': 600}], 'extra_recordings': [{'var': 'v', 'somadistance': 660, 'type': 'somadistance', 'name': 'dend1', 'seclist_name': 'apical', 'arbor_branch_index': 251, 'arbor_branch_index_with_replaced_axon': 123}, {'var': 'v', 'somadistance': 800, 'type': 'somadistance', 'name': 'dend2', 'seclist_name': 'apical', 'arbor_branch_index': 251, 'arbor_branch_index_with_replaced_axon': 123}]}, 'Step3': {'stimuli': [{'delay': 700, 'amp': 0.95, 'duration': 2000, 'totduration': 3000}, {'delay': 0, 'amp': -0.126, 'duration': 3000, 'totduration': 3000}]}, 'Step2': {'stimuli': [{'delay': 700, 'amp': 0.562, 'duration': 2000, 'totduration': 3000}, {'delay': 0, 'amp': -0.126, 'duration': 3000, 'totduration': 3000}]}, 'Step1': {'stimuli': [{'delay': 700, 'amp': 0.458, 'duration': 2000, 'totduration': 3000}, {'delay': 0, 'amp': -0.126, 'duration': 3000, 'totduration': 3000}]}}\n"
     ]
    }
   ],
   "source": [
    "proto_configs = json.load(open('config/protocols.json'))\n",
    "print(proto_configs)"
   ]
  },
  {
   "attachments": {},
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "And they can be automatically loaded"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 11,
   "metadata": {
    "collapsed": false,
    "jupyter": {
     "outputs_hidden": false
    }
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "bAP:\n",
      "  stimuli:\n",
      "    Square pulse amp 1.900000 delay 295.000000 duration 5.000000 totdur 600.000000 at ArbBranchRelLocation '(locset-def \"soma\" (location 0 0.5))'\n",
      "  recordings:\n",
      "    bAP.soma.v: v at ArbBranchRelLocation '(locset-def \"soma\" (location 0 0.5))'\n",
      "    bAP.dend1.v: v at ArbLocsetLocation (locset-def \"dend1\" (restrict (distal-translate (proximal (region \"apic\")) 660) (proximal-interval (distal (branch 123)))))\n",
      "    bAP.dend2.v: v at ArbLocsetLocation (locset-def \"dend2\" (restrict (distal-translate (proximal (region \"apic\")) 800) (proximal-interval (distal (branch 123)))))\n",
      "\n",
      "Step3:\n",
      "  stimuli:\n",
      "    Square pulse amp 0.950000 delay 700.000000 duration 2000.000000 totdur 3000.000000 at ArbBranchRelLocation '(locset-def \"soma\" (location 0 0.5))'\n",
      "    Square pulse amp -0.126000 delay 0.000000 duration 3000.000000 totdur 3000.000000 at ArbBranchRelLocation '(locset-def \"soma\" (location 0 0.5))'\n",
      "  recordings:\n",
      "    Step3.soma.v: v at ArbBranchRelLocation '(locset-def \"soma\" (location 0 0.5))'\n",
      "\n",
      "Step2:\n",
      "  stimuli:\n",
      "    Square pulse amp 0.562000 delay 700.000000 duration 2000.000000 totdur 3000.000000 at ArbBranchRelLocation '(locset-def \"soma\" (location 0 0.5))'\n",
      "    Square pulse amp -0.126000 delay 0.000000 duration 3000.000000 totdur 3000.000000 at ArbBranchRelLocation '(locset-def \"soma\" (location 0 0.5))'\n",
      "  recordings:\n",
      "    Step2.soma.v: v at ArbBranchRelLocation '(locset-def \"soma\" (location 0 0.5))'\n",
      "\n",
      "Step1:\n",
      "  stimuli:\n",
      "    Square pulse amp 0.458000 delay 700.000000 duration 2000.000000 totdur 3000.000000 at ArbBranchRelLocation '(locset-def \"soma\" (location 0 0.5))'\n",
      "    Square pulse amp -0.126000 delay 0.000000 duration 3000.000000 totdur 3000.000000 at ArbBranchRelLocation '(locset-def \"soma\" (location 0 0.5))'\n",
      "  recordings:\n",
      "    Step1.soma.v: v at ArbBranchRelLocation '(locset-def \"soma\" (location 0 0.5))'\n",
      "\n"
     ]
    }
   ],
   "source": [
    "import l5pc_evaluator\n",
    "fitness_protocols = l5pc_evaluator.define_protocols(do_replace_axon=morphology.do_replace_axon, sim='arb')\n",
    "print('\\n'.join('%s' % protocol for protocol in fitness_protocols.values()))"
   ]
  },
  {
   "attachments": {},
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## eFeatures"
   ]
  },
  {
   "attachments": {},
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "For every protocol we need to define which eFeatures will be used as objectives of the optimisation algorithm."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 12,
   "metadata": {
    "collapsed": false,
    "jupyter": {
     "outputs_hidden": false
    }
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "{ 'Step1': { 'soma': { 'AHP_depth_abs': [-60.3636, 2.3018],\n",
      "                       'AHP_depth_abs_slow': [-61.1513, 2.3385],\n",
      "                       'AHP_slow_time': [0.1599, 0.0483],\n",
      "                       'AP_height': [25.0141, 3.1463],\n",
      "                       'AP_width': [3.5312, 0.8592],\n",
      "                       'ISI_CV': [0.109, 0.1217],\n",
      "                       'adaptation_index2': [0.0047, 0.0514],\n",
      "                       'doublet_ISI': [62.75, 9.6667],\n",
      "                       'mean_frequency': [6, 1.2222],\n",
      "                       'time_to_first_spike': [27.25, 5.7222]}},\n",
      "  'Step2': { 'soma': { 'AHP_depth_abs': [-59.9055, 1.8329],\n",
      "                       'AHP_depth_abs_slow': [-60.2471, 1.8972],\n",
      "                       'AHP_slow_time': [0.1676, 0.0339],\n",
      "                       'AP_height': [27.1003, 3.1463],\n",
      "                       'AP_width': [2.7917, 0.7499],\n",
      "                       'ISI_CV': [0.0674, 0.075],\n",
      "                       'adaptation_index2': [0.005, 0.0067],\n",
      "                       'doublet_ISI': [44.0, 7.1327],\n",
      "                       'mean_frequency': [8.5, 0.9796],\n",
      "                       'time_to_first_spike': [19.75, 2.8776]}},\n",
      "  'Step3': { 'soma': { 'AHP_depth_abs': [-57.0905, 2.3427],\n",
      "                       'AHP_depth_abs_slow': [-61.1513, 2.3385],\n",
      "                       'AHP_slow_time': [0.1968, 0.0112],\n",
      "                       'AP_height': [19.7207, 3.7204],\n",
      "                       'AP_width': [3.5347, 0.8788],\n",
      "                       'ISI_CV': [0.0737, 0.0292],\n",
      "                       'adaptation_index2': [0.0055, 0.0015],\n",
      "                       'doublet_ISI': [22.75, 4.14],\n",
      "                       'mean_frequency': [17.5, 0.8],\n",
      "                       'time_to_first_spike': [10.5, 1.36]}},\n",
      "  'bAP': { 'dend1': {'AP_amplitude_from_voltagebase': [45, 10]},\n",
      "           'dend2': {'AP_amplitude_from_voltagebase': [36, 9.33]},\n",
      "           'soma': { 'AP_height': [25.0, 5.0],\n",
      "                     'AP_width': [2.0, 0.5],\n",
      "                     'Spikecount': [1.0, 0.01]}}}\n"
     ]
    }
   ],
   "source": [
    "feature_configs = json.load(open('config/features.json'))\n",
    "pp.pprint(feature_configs)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 13,
   "metadata": {
    "collapsed": false,
    "jupyter": {
     "outputs_hidden": false
    }
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "objectives:\n",
      "  ( AP_width for {'': 'bAP.soma.v'} with stim start 295 and end 600, exp mean 2.0 and std 0.5 and AP threshold override -20 )\n",
      "  ( AP_height for {'': 'bAP.soma.v'} with stim start 295 and end 600, exp mean 25.0 and std 5.0 and AP threshold override -20 )\n",
      "  ( Spikecount for {'': 'bAP.soma.v'} with stim start 295 and end 600, exp mean 1.0 and std 0.01 and AP threshold override -20 )\n",
      "  ( AP_amplitude_from_voltagebase for {'': 'bAP.dend1.v'} with stim start 295 and end 600, exp mean 45 and std 10 and AP threshold override -55 )\n",
      "  ( AP_amplitude_from_voltagebase for {'': 'bAP.dend2.v'} with stim start 295 and end 600, exp mean 36 and std 9.33 and AP threshold override -55 )\n",
      "  ( AP_height for {'': 'Step3.soma.v'} with stim start 700 and end 2700, exp mean 19.7207 and std 3.7204 and AP threshold override -20 )\n",
      "  ( AHP_slow_time for {'': 'Step3.soma.v'} with stim start 700 and end 2700, exp mean 0.1968 and std 0.0112 and AP threshold override -20 )\n",
      "  ( ISI_CV for {'': 'Step3.soma.v'} with stim start 700 and end 2700, exp mean 0.0737 and std 0.0292 and AP threshold override -20 )\n",
      "  ( doublet_ISI for {'': 'Step3.soma.v'} with stim start 700 and end 2700, exp mean 22.75 and std 4.14 and AP threshold override -20 )\n",
      "  ( adaptation_index2 for {'': 'Step3.soma.v'} with stim start 700 and end 2700, exp mean 0.0055 and std 0.0015 and AP threshold override -20 )\n",
      "  ( mean_frequency for {'': 'Step3.soma.v'} with stim start 700 and end 2700, exp mean 17.5 and std 0.8 and AP threshold override -20 )\n",
      "  ( AHP_depth_abs_slow for {'': 'Step3.soma.v'} with stim start 700 and end 2700, exp mean -61.1513 and std 2.3385 and AP threshold override -20 )\n",
      "  ( AP_width for {'': 'Step3.soma.v'} with stim start 700 and end 2700, exp mean 3.5347 and std 0.8788 and AP threshold override -20 )\n",
      "  ( time_to_first_spike for {'': 'Step3.soma.v'} with stim start 700 and end 2700, exp mean 10.5 and std 1.36 and AP threshold override -20 )\n",
      "  ( AHP_depth_abs for {'': 'Step3.soma.v'} with stim start 700 and end 2700, exp mean -57.0905 and std 2.3427 and AP threshold override -20 )\n",
      "  ( AP_height for {'': 'Step2.soma.v'} with stim start 700 and end 2700, exp mean 27.1003 and std 3.1463 and AP threshold override -20 )\n",
      "  ( AHP_slow_time for {'': 'Step2.soma.v'} with stim start 700 and end 2700, exp mean 0.1676 and std 0.0339 and AP threshold override -20 )\n",
      "  ( ISI_CV for {'': 'Step2.soma.v'} with stim start 700 and end 2700, exp mean 0.0674 and std 0.075 and AP threshold override -20 )\n",
      "  ( doublet_ISI for {'': 'Step2.soma.v'} with stim start 700 and end 2700, exp mean 44.0 and std 7.1327 and AP threshold override -20 )\n",
      "  ( adaptation_index2 for {'': 'Step2.soma.v'} with stim start 700 and end 2700, exp mean 0.005 and std 0.0067 and AP threshold override -20 )\n",
      "  ( mean_frequency for {'': 'Step2.soma.v'} with stim start 700 and end 2700, exp mean 8.5 and std 0.9796 and AP threshold override -20 )\n",
      "  ( AHP_depth_abs_slow for {'': 'Step2.soma.v'} with stim start 700 and end 2700, exp mean -60.2471 and std 1.8972 and AP threshold override -20 )\n",
      "  ( AP_width for {'': 'Step2.soma.v'} with stim start 700 and end 2700, exp mean 2.7917 and std 0.7499 and AP threshold override -20 )\n",
      "  ( time_to_first_spike for {'': 'Step2.soma.v'} with stim start 700 and end 2700, exp mean 19.75 and std 2.8776 and AP threshold override -20 )\n",
      "  ( AHP_depth_abs for {'': 'Step2.soma.v'} with stim start 700 and end 2700, exp mean -59.9055 and std 1.8329 and AP threshold override -20 )\n",
      "  ( AP_height for {'': 'Step1.soma.v'} with stim start 700 and end 2700, exp mean 25.0141 and std 3.1463 and AP threshold override -20 )\n",
      "  ( AHP_slow_time for {'': 'Step1.soma.v'} with stim start 700 and end 2700, exp mean 0.1599 and std 0.0483 and AP threshold override -20 )\n",
      "  ( ISI_CV for {'': 'Step1.soma.v'} with stim start 700 and end 2700, exp mean 0.109 and std 0.1217 and AP threshold override -20 )\n",
      "  ( doublet_ISI for {'': 'Step1.soma.v'} with stim start 700 and end 2700, exp mean 62.75 and std 9.6667 and AP threshold override -20 )\n",
      "  ( adaptation_index2 for {'': 'Step1.soma.v'} with stim start 700 and end 2700, exp mean 0.0047 and std 0.0514 and AP threshold override -20 )\n",
      "  ( mean_frequency for {'': 'Step1.soma.v'} with stim start 700 and end 2700, exp mean 6 and std 1.2222 and AP threshold override -20 )\n",
      "  ( AHP_depth_abs_slow for {'': 'Step1.soma.v'} with stim start 700 and end 2700, exp mean -61.1513 and std 2.3385 and AP threshold override -20 )\n",
      "  ( AP_width for {'': 'Step1.soma.v'} with stim start 700 and end 2700, exp mean 3.5312 and std 0.8592 and AP threshold override -20 )\n",
      "  ( time_to_first_spike for {'': 'Step1.soma.v'} with stim start 700 and end 2700, exp mean 27.25 and std 5.7222 and AP threshold override -20 )\n",
      "  ( AHP_depth_abs for {'': 'Step1.soma.v'} with stim start 700 and end 2700, exp mean -60.3636 and std 2.3018 and AP threshold override -20 )\n"
     ]
    }
   ],
   "source": [
    "fitness_calculator = l5pc_evaluator.define_fitness_calculator(fitness_protocols)\n",
    "print(fitness_calculator)"
   ]
  },
  {
   "attachments": {},
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Simulator"
   ]
  },
  {
   "attachments": {},
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "We need to define which simulator we will use. In this case it will be Arbor, i.e. the ArbSimulator class"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 14,
   "metadata": {
    "collapsed": false,
    "jupyter": {
     "outputs_hidden": false
    }
   },
   "outputs": [],
   "source": [
    "sim = ephys.simulators.ArbSimulator()"
   ]
  },
  {
   "attachments": {},
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Evaluator"
   ]
  },
  {
   "attachments": {},
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "With all the components defined above we can build a cell evaluator"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 15,
   "metadata": {
    "collapsed": false,
    "jupyter": {
     "outputs_hidden": false
    }
   },
   "outputs": [],
   "source": [
    "evaluator = ephys.evaluators.CellEvaluator(                                          \n",
    "        cell_model=l5pc_cell,                                                       \n",
    "        param_names=param_names,                                                    \n",
    "        fitness_protocols=fitness_protocols,                                        \n",
    "        fitness_calculator=fitness_calculator,                                      \n",
    "        sim=sim)  "
   ]
  },
  {
   "attachments": {},
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "This evaluator can be used to run the protocols. The original parameter values for the Markram et al. 2015 L5PC model are:"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 16,
   "metadata": {},
   "outputs": [],
   "source": [
    "release_params = {\n",
    "    'gNaTs2_tbar_NaTs2_t.apical': 0.026145,\n",
    "    'gSKv3_1bar_SKv3_1.apical': 0.004226,\n",
    "    'gImbar_Im.apical': 0.000143,\n",
    "    'gNaTa_tbar_NaTa_t.axonal': 3.137968,\n",
    "    'gK_Tstbar_K_Tst.axonal': 0.089259,\n",
    "    'gamma_CaDynamics_E2.axonal': 0.002910,\n",
    "    'gNap_Et2bar_Nap_Et2.axonal': 0.006827,\n",
    "    'gSK_E2bar_SK_E2.axonal': 0.007104,\n",
    "    'gCa_HVAbar_Ca_HVA.axonal': 0.000990,\n",
    "    'gK_Pstbar_K_Pst.axonal': 0.973538,\n",
    "    'gSKv3_1bar_SKv3_1.axonal': 1.021945,\n",
    "    'decay_CaDynamics_E2.axonal': 287.198731,\n",
    "    'gCa_LVAstbar_Ca_LVAst.axonal': 0.008752,\n",
    "    'gamma_CaDynamics_E2.somatic': 0.000609,\n",
    "    'gSKv3_1bar_SKv3_1.somatic': 0.303472,\n",
    "    'gSK_E2bar_SK_E2.somatic': 0.008407,\n",
    "    'gCa_HVAbar_Ca_HVA.somatic': 0.000994,\n",
    "    'gNaTs2_tbar_NaTs2_t.somatic': 0.983955,\n",
    "    'decay_CaDynamics_E2.somatic': 210.485284,\n",
    "    'gCa_LVAstbar_Ca_LVAst.somatic': 0.000333\n",
    "}\n"
   ]
  },
  {
   "attachments": {},
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Running the responses is as easy as passing the protocols and parameters to the evaluator. (The line below will take some time to execute)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 17,
   "metadata": {
    "collapsed": false,
    "jupyter": {
     "outputs_hidden": false
    }
   },
   "outputs": [],
   "source": [
    "release_responses = evaluator.run_protocols(protocols=fitness_protocols.values(), param_values=release_params)"
   ]
  },
  {
   "attachments": {},
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "We can now plot all the responses"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 18,
   "metadata": {
    "collapsed": false,
    "jupyter": {
     "outputs_hidden": false
    }
   },
   "outputs": [
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 720x720 with 6 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "def plot_responses(responses):\n",
    "    fig, axes = plt.subplots(len(responses), figsize=(10,10))\n",
    "    for index, (resp_name, response) in enumerate(sorted(responses.items())):\n",
    "        axes[index].plot(response['time'], response['voltage'], label=resp_name)\n",
    "        axes[index].set_title(resp_name)\n",
    "    fig.tight_layout()\n",
    "plot_responses(release_responses)"
   ]
  },
  {
   "attachments": {},
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Running an optimisation of the parameters now has become very easy. \n",
    "Of course running the L5PC optimisation will require quite some computing resources. \n",
    "\n",
    "To show a proof-of-concept, we will only run 2 generations, with 2 offspring individuals per generations.\n",
    "If you want to run all full optimisation, you should run for 100 generations with an offspring size of 100 individuals. "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 19,
   "metadata": {
    "collapsed": false,
    "jupyter": {
     "outputs_hidden": false
    }
   },
   "outputs": [],
   "source": [
    "opt = bpopt.optimisations.DEAPOptimisation(                                     \n",
    "    evaluator=evaluator,                                                            \n",
    "    offspring_size=2) \n",
    "final_pop, halloffame, log, hist = opt.run(max_ngen=2, cp_filename='checkpoints/checkpoint.pkl')"
   ]
  },
  {
   "attachments": {},
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "The first individual in the hall of fame will contain the best solution found."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 20,
   "metadata": {
    "collapsed": false,
    "jupyter": {
     "outputs_hidden": false
    }
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "[0.001017834439738432, 0.021656498911739864, 0.0009391491627785106, 1.5248169507528497, 0.8663975885224535, 0.4221165755827173, 0.0029040787574867947, 0.022169166627303505, 0.8757751873011441, 0.0004958122413818507, 0.002330844502575726, 0.011927893806278723, 234.40541659093483, 0.4596034657377336, 0.28978161459048557, 0.0021489705265908877, 0.0008375779756625729, 0.005564543226524335, 0.03229357096515606, 202.18814057682334]\n"
     ]
    }
   ],
   "source": [
    "print(halloffame[0])"
   ]
  },
  {
   "attachments": {},
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "These are the raw parameter values. \n",
    "The evaluator object can convert this in a dictionary, so that we can see the parameter names corresponding to these values."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 21,
   "metadata": {
    "collapsed": false,
    "jupyter": {
     "outputs_hidden": false
    }
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "{ 'decay_CaDynamics_E2.axonal': 234.40541659093483,\n",
      "  'decay_CaDynamics_E2.somatic': 202.18814057682334,\n",
      "  'gCa_HVAbar_Ca_HVA.axonal': 0.0004958122413818507,\n",
      "  'gCa_HVAbar_Ca_HVA.somatic': 0.0008375779756625729,\n",
      "  'gCa_LVAstbar_Ca_LVAst.axonal': 0.002330844502575726,\n",
      "  'gCa_LVAstbar_Ca_LVAst.somatic': 0.005564543226524335,\n",
      "  'gImbar_Im.apical': 0.0009391491627785106,\n",
      "  'gK_Pstbar_K_Pst.axonal': 0.4221165755827173,\n",
      "  'gK_Tstbar_K_Tst.axonal': 0.0029040787574867947,\n",
      "  'gNaTa_tbar_NaTa_t.axonal': 1.5248169507528497,\n",
      "  'gNaTs2_tbar_NaTs2_t.apical': 0.001017834439738432,\n",
      "  'gNaTs2_tbar_NaTs2_t.somatic': 0.4596034657377336,\n",
      "  'gNap_Et2bar_Nap_Et2.axonal': 0.8663975885224535,\n",
      "  'gSK_E2bar_SK_E2.axonal': 0.022169166627303505,\n",
      "  'gSK_E2bar_SK_E2.somatic': 0.0021489705265908877,\n",
      "  'gSKv3_1bar_SKv3_1.apical': 0.021656498911739864,\n",
      "  'gSKv3_1bar_SKv3_1.axonal': 0.8757751873011441,\n",
      "  'gSKv3_1bar_SKv3_1.somatic': 0.28978161459048557,\n",
      "  'gamma_CaDynamics_E2.axonal': 0.011927893806278723,\n",
      "  'gamma_CaDynamics_E2.somatic': 0.03229357096515606}\n",
      "None\n"
     ]
    }
   ],
   "source": [
    "best_params = evaluator.param_dict(halloffame[0])\n",
    "print(pp.pprint(best_params))"
   ]
  },
  {
   "attachments": {},
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Then we can run the fitness protocols on the model with these parameter values"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 22,
   "metadata": {
    "collapsed": false,
    "jupyter": {
     "outputs_hidden": false
    }
   },
   "outputs": [],
   "source": [
    "best_responses = evaluator.run_protocols(protocols=fitness_protocols.values(), param_values=best_params)"
   ]
  },
  {
   "attachments": {},
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "And then we can also plot these responses. \n",
    "\n",
    "When you ran the above optimisation with only 2 individuals and 2 generations, this 'best' model will of course be very low quality."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 23,
   "metadata": {
    "collapsed": false,
    "jupyter": {
     "outputs_hidden": false
    }
   },
   "outputs": [
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 720x720 with 6 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "plot_responses(best_responses)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": []
  }
 ],
 "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": "3.8.10"
  },
  "vscode": {
   "interpreter": {
    "hash": "581988038cf9ce8838e7faf3da7c29f4ff88d898cd43cb17e0086e389d8deda2"
   }
  }
 },
 "nbformat": 4,
 "nbformat_minor": 4
}