Gromacs_inputs.ipynb

{
  "nbformat": 4,
  "nbformat_minor": 0,
  "metadata": {
    "accelerator": "GPU",
    "colab": {
      "provenance": [],
      "include_colab_link": true
    },
    "kernelspec": {
      "display_name": "Python 3",
      "name": "python3"
    }
  },
  "cells": [
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "view-in-github",
        "colab_type": "text"
      },
      "source": [
        "<a href=\"https://colab.research.google.com/github/pablo-arantes/making-it-rain/blob/main/Gromacs_inputs.ipynb\" target=\"_parent\"><img src=\"https://colab.research.google.com/assets/colab-badge.svg\" alt=\"Open In Colab\"/></a>"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "pj2BWZxUDbEE"
      },
      "source": [
        "# **Hello there!**\n",
        "\n",
        "This is a Jupyter notebook for running Molecular Dynamics (MD) simulations using OpenMM engine and GROMACS biomolecular simulation package inputs. This notebook is a supplementary material of the paper \"***Making it rain: Cloud-based molecular simulations for everyone***\" ([link here](https://doi.org/10.1021/acs.jcim.1c00998)) and we encourage you to read it before using this pipeline.\n",
        "\n",
        "The main goal of this notebook is to demonstrate how to harness the power of cloud-computing to run microsecond-long MD simulations in a cheap and yet feasible fashion.\n",
        "\n",
        "---\n",
        "\n",
        " **This notebook is NOT a standard protocol for MD simulations!** It is just simple MD pipeline illustrating each step of a simulation protocol.\n",
        "\n",
        "---\n",
        "**Bugs**\n",
        "- If you encounter any bugs, please report the issue to https://github.com/pablo-arantes/making-it-rain/issues\n",
        "\n",
        "**Acknowledgments**\n",
        "- We would like to thank the OpenMM team for developing an excellent and open source engine.\n",
        "\n",
        "- A Making-it-rain by **Pablo R. Arantes** ([@pablitoarantes](https://twitter.com/pablitoarantes)), **Marcelo D. Polêto** ([@mdpoleto](https://twitter.com/mdpoleto)), **Conrado Pedebos** ([@ConradoPedebos](https://twitter.com/ConradoPedebos)) and **Rodrigo Ligabue-Braun** ([@ligabue_braun](https://twitter.com/ligabue_braun)).\n",
        "\n",
        "\n",
        "- Also, credit to [David Koes](https://github.com/dkoes) for his awesome [py3Dmol](https://3dmol.csb.pitt.edu/) plugin.\n",
        "\n",
        "- For related notebooks see: [Making-it-rain](https://github.com/pablo-arantes/making-it-rain)"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "hoyY6XonD1UX"
      },
      "source": [
        "# **Introduction**\n",
        "\n",
        "In general, MD simulations rely on 1) a set of atomic coordinates of all atoms on a simulation box and 2) a set of force field parameters that describes the interaction energies between atoms.\n",
        "\n",
        "In terms of GROMACS inputs, we will need:\n",
        "\n",
        "*   A .gro file containing a set of atomic coordinates.\n",
        "*   The respective .top file containing the system topology.\n",
        "\n",
        "In this notebook, we will simulate PDB 1AKI, a hen egg-white lysozyme. To build our simulation box, we used GROMACS package (https://www.gromacs.org/). You can download the input files examples from [here](https://github.com/pablo-arantes/making-it-rain/tree/main/GROMACS_INPUTS);\n",
        "## ---\n",
        "\n",
        "\n",
        "\n",
        "\n",
        "\n"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "Lh96y6mGFY1D"
      },
      "source": [
        "---\n",
        "---\n",
        "# **Setting the environment for MD calculation**\n",
        "\n",
        "Firstly, we need to install all necessary libraries and packages for our simulation. The main packages we will be installing are:\n",
        "\n",
        "1.    Anaconda (https://docs.conda.io/en/latest/miniconda.html)\n",
        "2.    OpenMM (https://openmm.org/)\n",
        "3.    PyTraj (https://amber-md.github.io/pytraj/latest/index.html)\n",
        "4.    py3Dmol (https://pypi.org/project/py3Dmol/)\n",
        "5.    Numpy (https://numpy.org/)\n",
        "6.    Matplotlib (https://matplotlib.org/)\n",
        "7.    AmberTools (https://ambermd.org/AmberTools.php)"
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "#@title **Install Conda Colab**\n",
        "#@markdown It will restart the kernel (session), don't worry.\n",
        "!pip install -q condacolab\n",
        "import condacolab\n",
        "condacolab.install()"
      ],
      "metadata": {
        "id": "_juUn0IkFM7F",
        "cellView": "form"
      },
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "source": [
        "#@title **Install dependencies**\n",
        "#@markdown It will take a few minutes, please, drink a coffee and wait. ;-)\n",
        "# install dependencies\n",
        "import subprocess\n",
        "import sys\n",
        "subprocess.run(\"rm -rf /usr/local/conda-meta/pinned\", shell=True)\n",
        "subprocess.run(\"mamba install -c conda-forge ambertools -y\", shell=True)\n",
        "import pytraj as pt\n",
        "subprocess.run(\"pip -q install py3Dmol\", shell=True)\n",
        "subprocess.run(\"pip install git+https://github.com/pablo-arantes/biopandas\", shell=True)\n",
        "subprocess.run(\"mamba install openmm=7.7.0\", shell=True)\n",
        "subprocess.run(\"pip install --upgrade MDAnalysis==2.4.2\", shell=True)\n",
        "\n",
        "#load dependencies\n",
        "import sys\n",
        "from biopandas.pdb import PandasPdb\n",
        "import openmm as mm\n",
        "from openmm import *\n",
        "from openmm.app import *\n",
        "from openmm.unit import *\n",
        "import os\n",
        "import urllib.request\n",
        "import numpy as np\n",
        "import MDAnalysis as mda\n",
        "import py3Dmol\n",
        "import pytraj as pt\n",
        "import platform\n",
        "import scipy.cluster.hierarchy\n",
        "from scipy.spatial.distance import squareform\n",
        "import scipy.stats as stats\n",
        "import matplotlib.pyplot as plt\n",
        "import pandas as pd\n",
        "from scipy.interpolate import griddata\n",
        "import seaborn as sb\n",
        "from statistics import mean, stdev\n",
        "from pytraj import matrix\n",
        "from matplotlib import colors\n",
        "from IPython.display import set_matplotlib_formats"
      ],
      "metadata": {
        "cellView": "form",
        "id": "aY1ghAqRFN34"
      },
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "fDQnAKJLFxtt"
      },
      "source": [
        "## Using Google Drive to store simulation data\n",
        "\n",
        "Google Colab does not allow users to keep data on their computing nodes. However, we can use Google Drive to read, write, and store our simulations files. Therefore, we suggest to you to:\n",
        "\n",
        "1.   Create a folder in your own Google Drive and copy the necessary input files there.\n",
        "2.   Copy the path of your created directory. We will use it below."
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "Lm7Akepv_vl-",
        "cellView": "form"
      },
      "source": [
        "#@title ### **Import Google Drive**\n",
        "#@markdown Click in the \"Run\" buttom to make your Google Drive accessible.\n",
        "from google.colab import drive\n",
        "\n",
        "drive.flush_and_unmount()\n",
        "drive.mount('/content/drive', force_remount=True)"
      ],
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "lOKg9eH_ueRn",
        "cellView": "form"
      },
      "source": [
        "#@title **Check if you correctly allocated GPU nodes**\n",
        "\n",
        "gpu_info = !nvidia-smi\n",
        "gpu_info = '\\n'.join(gpu_info)\n",
        "if gpu_info.find('failed') >= 0:\n",
        "  print('Select the Runtime > \"Change runtime type\" menu to enable a GPU accelerator, ')\n",
        "  print('and then re-execute this cell.')\n",
        "else:\n",
        "  print(gpu_info)"
      ],
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "BZwl66HTGI7v"
      },
      "source": [
        "---\n",
        "# **Loading the necessary input files**\n",
        "\n",
        "At this point, we should have all libraries and dependencies installed and all necessary input files already at your Google Drive folder.\n",
        "\n",
        "Below, you should provide the names of all input files (**gro and top**) and the pathway of your Google Drive folder containing them.\n",
        "\n",
        "**Attention:** The OpenMM engine accepts **AMBER, CHARMM and OPLS** force fields in GROMACS format. The GROMOS force field is not accepted.\n",
        "\n",
        "**Please, don't use spaces in the files and folders names, i.e., protein_topol.top, MyDrive/GROMACS and so on.**"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "cellView": "form",
        "id": "0sfrVUYeXhtp"
      },
      "source": [
        "import warnings\n",
        "warnings.filterwarnings('ignore')\n",
        "\n",
        "#@title **Please, provide the necessary input files below:**\n",
        "TOP_filename = 'topol.top' #@param {type:\"string\"}\n",
        "GRO_filename = '1AKI_solv_ions.gro' #@param {type:\"string\"}\n",
        "\n",
        "Google_Drive_Path = '/content/drive/MyDrive/GROMACS' #@param {type:\"string\"}\n",
        "workDir = Google_Drive_Path\n",
        "\n",
        "#@markdown If you have your own force field folder in  your Google Drive path, please, select yes and then type the folder name (**THIS STEP IS NOT MANDATORY**):\n",
        "\n",
        "Force_field_folder = 'No' #@param [\"Yes\", \"No\"]\n",
        "Foldername = '' #@param {type:\"string\"}\n",
        "\n",
        "if Force_field_folder == 'No':\n",
        "  foldername = 'top'\n",
        "  top_folder = os.path.join(workDir, str(foldername))\n",
        "  top_folder_check = os.path.exists(top_folder)\n",
        "  if top_folder_check == False:\n",
        "    !npx degit https://github.com/pablo-arantes/Making-it-rain/ temp 2> /dev/null\n",
        "    cp_command = \"cp -r temp/top \" + workDir\n",
        "    original_stdout = sys.stdout # Save a reference to the original standard output\n",
        "    with open('cp.sh', 'w') as f:\n",
        "        sys.stdout = f # Change the standard output to the file we created.\n",
        "        print(cp_command)\n",
        "        sys.stdout = original_stdout # Reset the standard output to its original value\n",
        "    !chmod 700 cp.sh 2>&1 1>/dev/null\n",
        "    !bash cp.sh 2>&1 1>/dev/null\n",
        "    !rm -r temp cp.sh\n",
        "else:\n",
        "  foldername = Foldername\n",
        "  top_folder = os.path.join(workDir, str(foldername))\n",
        "\n",
        "top = os.path.join(workDir, str(TOP_filename))\n",
        "gro = os.path.join(workDir, str(GRO_filename))\n",
        "\n",
        "universe = mda.Universe(gro)\n",
        "with mda.Writer(os.path.join(workDir, \"SYS.pdb\")) as pdb:\n",
        "    pdb.write(universe)\n",
        "\n",
        "PDB_filename = 'SYS.pdb'\n",
        "pdb = os.path.join(workDir, str(PDB_filename))\n",
        "\n",
        "pdb_gromacs = os.path.exists(pdb)\n",
        "top_gromacs = os.path.exists(top)\n",
        "gro_gromacs = os.path.exists(gro)\n",
        "top_folder_check = os.path.exists(top_folder)\n",
        "\n",
        "if pdb_gromacs == True and top_gromacs == True and gro_gromacs == True and top_folder_check == True:\n",
        "    print(\"Files loaded succesfully! ;-)\")\n",
        "else:\n",
        "    print(\"ERROR! Check your input names and Google Drive path\")"
      ],
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "C8kKR7bpI86W"
      },
      "source": [
        "## Let's take a look on our simulation box:"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "vmQ27nZLssjv",
        "cellView": "form"
      },
      "source": [
        "#@title **Show 3D structure**\n",
        "import warnings\n",
        "warnings.filterwarnings('ignore')\n",
        "import py3Dmol\n",
        "\n",
        "color = \"gray\" #@param [\"gray\", \"rainbow\"]\n",
        "show_sidechains = False #@param {type:\"boolean\"}\n",
        "show_mainchains = False #@param {type:\"boolean\"}\n",
        "show_box = True #@param {type:\"boolean\"}\n",
        "box_opacity = 0.6 #@param {type:\"slider\", min:0, max:1, step:0.1}\n",
        "\n",
        "\n",
        "def show_pdb(show_sidechains=False, show_mainchains=False, show_box = False, color=\"rainbow\"):\n",
        "  view = py3Dmol.view(js='https://3dmol.org/build/3Dmol.js',)\n",
        "  view.addModel(open(pdb,'r').read(),'pdb')\n",
        "\n",
        "  if color == \"gray\":\n",
        "    view.setStyle({'cartoon':{}})\n",
        "  elif color == \"rainbow\":\n",
        "    view.setStyle({'cartoon': {'color':'spectrum'}})\n",
        "\n",
        "  if show_sidechains:\n",
        "    BB = ['C','O','N']\n",
        "    view.addStyle({'and':[{'resn':[\"GLY\",\"PRO\"],'invert':True},{'atom':BB,'invert':True}]},\n",
        "                        {'stick':{'colorscheme':f\"WhiteCarbon\",'radius':0.3}})\n",
        "    view.addStyle({'and':[{'resn':\"GLY\"},{'atom':'CA'}]},\n",
        "                        {'sphere':{'colorscheme':f\"WhiteCarbon\",'radius':0.3}})\n",
        "    view.addStyle({'and':[{'resn':\"PRO\"},{'atom':['C','O'],'invert':True}]},\n",
        "                        {'stick':{'colorscheme':f\"WhiteCarbon\",'radius':0.3}})\n",
        "  if show_mainchains:\n",
        "    BB = ['C','O','N','CA']\n",
        "    view.addStyle({'atom':BB},{'stick':{'colorscheme':f\"WhiteCarbon\",'radius':0.3}})\n",
        "\n",
        "  if show_box:\n",
        "    view.addSurface(py3Dmol.SAS, {'opacity': box_opacity, 'color':'white'})\n",
        "\n",
        "  view.zoomTo()\n",
        "  return view\n",
        "\n",
        "\n",
        "show_pdb(show_sidechains, show_mainchains, show_box, color).show()"
      ],
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "n85MrAO7M7uQ"
      },
      "source": [
        "---\n",
        "---\n",
        "# **Equilibrating the simulation box**\n",
        "\n",
        "Proper MD equilibration protocol is designed to equilibrate both temperature and pressure throughout the simulation box while preserving the protein experimental conformation. In addition, we also allow the solvent to accomodate around the protein, creating proper solvation layers.\n",
        "\n",
        "Below, we will set up the MD equilibration parameters, such as temperature, pressure and the desired simulation time. We will define the force constant used to restraint protein heavy-atoms in place and the frequency at which we want to save atomic coordinates in a trajectory file (.dcd).\n",
        "\n",
        "After you are done, you can run the next 2 cells to equilibrate your system."
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "8x9Qp_dbr9HP",
        "cellView": "form"
      },
      "source": [
        "#@title ### **Parameters for MD Equilibration protocol:**\n",
        "\n",
        "# remove whitespaces\n",
        "Jobname = '1aki_equil' #@param {type:\"string\"}\n",
        "\n",
        "Minimization_steps = \"1000\" #@param [\"1000\", \"5000\", \"10000\", \"20000\", \"50000\", \"100000\"]\n",
        "\n",
        "#@markdown Simulation time (in nanoseconds) and integration time (in femtoseconds):\n",
        "Time = \"2\" #@param {type:\"string\"}\n",
        "stride_time_eq = Time\n",
        "Integration_timestep = \"2\" #@param [\"0.5\", \"1\", \"2\", \"3\", \"4\"]\n",
        "dt_eq = Integration_timestep\n",
        "\n",
        "#@markdown Temperature (in Kelvin) and Pressure (in bar)\n",
        "Temperature = 298 #@param {type:\"string\"}\n",
        "temperature_eq = Temperature\n",
        "Pressure = 1 #@param {type:\"string\"}\n",
        "pressure_eq = Pressure\n",
        "\n",
        "#@markdown Position restraints force constant (in kJ/mol):\n",
        "Force_constant = 800 #@param {type:\"slider\", min:0, max:2000, step:100}\n",
        "\n",
        "#@markdown Frequency to write the trajectory file (in picoseconds):\n",
        "\n",
        "Write_the_trajectory = \"10\" #@param [\"10\", \"100\", \"200\", \"500\", \"1000\"]\n",
        "write_the_trajectory_eq = Write_the_trajectory\n",
        "#@markdown Frequency to write the log file (in picoseconds):\n",
        "\n",
        "Write_the_log = \"10\" #@param [\"10\", \"100\", \"200\", \"500\", \"1000\"]\n",
        "write_the_log_eq = Write_the_log\n",
        "\n",
        "\n",
        "#@markdown ---\n"
      ],
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "zoamR9iynphz",
        "cellView": "form"
      },
      "source": [
        "#@title **Runs an Equilibration MD simulation (NPT ensemble)**\n",
        "#@markdown Now, let's equilibrate our system!\n",
        "\n",
        "###########################################\n",
        "import openmm as mm\n",
        "from openmm import *\n",
        "from openmm.app import *\n",
        "from openmm.unit import *\n",
        "import pytraj as pt\n",
        "\n",
        "from sys import stdout, exit, stderr\n",
        "import os, math, fnmatch\n",
        "\n",
        "#############################################\n",
        "# Defining MD simulation parameters\n",
        "\n",
        "jobname = os.path.join(workDir, Jobname)\n",
        "coordinatefile = os.path.join(workDir, str(GRO_filename))\n",
        "pdbfile = os.path.join(workDir, str(PDB_filename))\n",
        "topologyfile = os.path.join(workDir, str(TOP_filename))\n",
        "\n",
        "time_ps = float(Time)*1000\n",
        "simulation_time = float(time_ps)*picosecond\t\t# in ps\n",
        "dt = int(dt_eq)*femtosecond\n",
        "temperature = float(temperature_eq)*kelvin\n",
        "savcrd_freq = int(write_the_trajectory_eq)*picosecond\n",
        "print_freq  = int(write_the_log_eq)*picosecond\n",
        "\n",
        "pressure\t= float(pressure_eq)*bar\n",
        "\n",
        "restraint_fc = int(Force_constant) # kJ/mol\n",
        "\n",
        "nsteps  = int(simulation_time.value_in_unit(picosecond)/dt.value_in_unit(picosecond))\n",
        "nprint  = int(print_freq.value_in_unit(picosecond)/dt.value_in_unit(picosecond))\n",
        "nsavcrd = int(savcrd_freq.value_in_unit(picosecond)/dt.value_in_unit(picosecond))\n",
        "\n",
        "#############################################\n",
        "# Defining functions to use below:\n",
        "def backup_old_log(pattern, string):\n",
        "\tresult = []\n",
        "\tfor root, dirs, files in os.walk(\"./\"):\n",
        "\t\tfor name in files:\n",
        "\t\t\tif fnmatch.fnmatch(name, pattern):\n",
        "\n",
        "\t\t\t\ttry:\n",
        "\t\t\t\t\tnumber = int(name[-2])\n",
        "\t\t\t\t\tavail = isinstance(number, int)\n",
        "\t\t\t\t\t#print(name,avail)\n",
        "\t\t\t\t\tif avail == True:\n",
        "\t\t\t\t\t\tresult.append(number)\n",
        "\t\t\t\texcept:\n",
        "\t\t\t\t\tpass\n",
        "\n",
        "\tif len(result) > 0:\n",
        "\t\tmaxnumber = max(result)\n",
        "\telse:\n",
        "\t\tmaxnumber = 0\n",
        "\n",
        "\tbackup_file = \"\\#\" + string + \".\" + str(maxnumber + 1) + \"#\"\n",
        "\tos.system(\"mv \" + string + \" \" + backup_file)\n",
        "\treturn backup_file\n",
        "\n",
        "def restraints(system, crd, fc, restraint_array):\n",
        "\n",
        "\tboxlx = system.getDefaultPeriodicBoxVectors()[0][0].value_in_unit(nanometers)\n",
        "\tboxly = system.getDefaultPeriodicBoxVectors()[1][1].value_in_unit(nanometers)\n",
        "\tboxlz = system.getDefaultPeriodicBoxVectors()[2][2].value_in_unit(nanometers)\n",
        "\n",
        "\tif fc > 0:\n",
        "\t\t# positional restraints for all heavy-atoms\n",
        "\t\tposresPROT = CustomExternalForce('k*periodicdistance(x, y, z, x0, y0, z0)^2;')\n",
        "\t\tposresPROT.addPerParticleParameter('k')\n",
        "\t\tposresPROT.addPerParticleParameter('x0')\n",
        "\t\tposresPROT.addPerParticleParameter('y0')\n",
        "\t\tposresPROT.addPerParticleParameter('z0')\n",
        "\n",
        "\t\tfor atom1 in restraint_array:\n",
        "\t\t\tatom1 = int(atom1)\n",
        "\n",
        "\t\t\txpos  = crd.positions[atom1].value_in_unit(nanometers)[0]\n",
        "\t\t\typos  = crd.positions[atom1].value_in_unit(nanometers)[1]\n",
        "\t\t\tzpos  = crd.positions[atom1].value_in_unit(nanometers)[2]\n",
        "\n",
        "\t\t\tposresPROT.addParticle(atom1, [fc, xpos, ypos, zpos])\n",
        "\n",
        "\t\tsystem.addForce(posresPROT)\n",
        "\n",
        "\treturn system\n",
        "##############################################\n",
        "\n",
        "#############################################\n",
        "print(\"\\n> Simulation details:\\n\")\n",
        "print(\"\\tJob name = \" + jobname)\n",
        "print(\"\\tCoordinate file = \" + str(coordinatefile))\n",
        "print(\"\\tPDB file = \" + str(pdbfile))\n",
        "print(\"\\tTopology file = \" + str(topologyfile))\n",
        "\n",
        "print(\"\\n\\tSimulation_time = \" + str(simulation_time))\n",
        "print(\"\\tIntegration timestep = \" + str(dt))\n",
        "print(\"\\tTotal number of steps = \" +  str(nsteps))\n",
        "\n",
        "print(\"\\n\\tSave coordinates each \" + str(savcrd_freq))\n",
        "print(\"\\tPrint in log file each \" + str(print_freq))\n",
        "\n",
        "print(\"\\n\\tTemperature = \" + str(temperature))\n",
        "print(\"\\tPressure = \" + str(pressure))\n",
        "#############################################\n",
        "\n",
        "print(\"\\n> Setting the system:\\n\")\n",
        "\n",
        "print(\"\\t- Reading topology and structure file...\")\n",
        "gro = GromacsGroFile(coordinatefile)\n",
        "top = GromacsTopFile(topologyfile, periodicBoxVectors=gro.getPeriodicBoxVectors(),\n",
        "        includeDir= os.path.join(workDir, str(foldername)))\n",
        "\n",
        "print(\"\\t- Creating system and setting parameters...\")\n",
        "nonbondedMethod = PME\n",
        "nonbondedCutoff = 1.0*nanometers\n",
        "ewaldErrorTolerance = 0.0005\n",
        "constraints = HBonds\n",
        "rigidWater = True\n",
        "constraintTolerance = 0.000001\n",
        "friction = 1.0\n",
        "system = top.createSystem(nonbondedMethod=nonbondedMethod, nonbondedCutoff=nonbondedCutoff,\n",
        "                           constraints=constraints, rigidWater=rigidWater, ewaldErrorTolerance=ewaldErrorTolerance)\n",
        "\n",
        "print(\"\\t- Applying restraints. Force Constant = \" + str(Force_constant) + \"kJ/mol\")\n",
        "pt_system = pt.iterload(coordinatefile, pdb)\n",
        "pt_topology = pt_system.top\n",
        "restraint_array = pt.select_atoms('!(:SOL) & !(:NA) & !(:CL) & !(:MG) & !(:K)', pt_topology)\n",
        "\n",
        "system = restraints(system, gro, restraint_fc, restraint_array)\n",
        "\n",
        "print(\"\\t- Setting barostat...\")\n",
        "system.addForce(MonteCarloBarostat(pressure, temperature))\n",
        "\n",
        "print(\"\\t- Setting integrator...\")\n",
        "integrator = LangevinIntegrator(temperature, friction, dt)\n",
        "integrator.setConstraintTolerance(constraintTolerance)\n",
        "simulation = Simulation(top.topology, system, integrator)\n",
        "simulation.context.setPositions(gro.positions)\n",
        "# if gro.boxVectors is not None:\n",
        "#     simulation.context.setPeriodicBoxVectors(topologyfile, periodicBoxVectors=gro.getPeriodicBoxVectors())\n",
        "\n",
        "print(\"\\t- Energy minimization: \" + str(Minimization_steps) + \" steps\")\n",
        "simulation.minimizeEnergy(tolerance=10*kilojoule/mole, maxIterations=int(Minimization_steps))\n",
        "print(\"\\t-> Potential Energy = \" + str(simulation.context.getState(getEnergy=True).getPotentialEnergy()))\n",
        "\n",
        "print(\"\\t- Setting initial velocities...\")\n",
        "simulation.context.setVelocitiesToTemperature(temperature)\n",
        "\n",
        "#############################################\n",
        "# Running Equilibration on NPT ensemble\n",
        "\n",
        "dcd_file = jobname + \".dcd\"\n",
        "log_file = jobname + \".log\"\n",
        "rst_file = jobname + \".rst\"\n",
        "prv_rst_file = jobname + \".rst\"\n",
        "pdb_file = jobname + \".pdb\"\n",
        "\n",
        "# Creating a trajectory file and reporters\n",
        "dcd = DCDReporter(dcd_file, nsavcrd)\n",
        "firstdcdstep = (nsteps) + nsavcrd\n",
        "dcd._dcd = DCDFile(dcd._out, simulation.topology, simulation.integrator.getStepSize(), firstdcdstep, nsavcrd) # charmm doesn't like first step to be 0\n",
        "\n",
        "simulation.reporters.append(dcd)\n",
        "simulation.reporters.append(StateDataReporter(stdout, nprint, step=True, speed=True, progress=True, totalSteps=nsteps, remainingTime=True, separator='\\t\\t'))\n",
        "simulation.reporters.append(StateDataReporter(log_file, nprint, step=True, kineticEnergy=True, potentialEnergy=True, totalEnergy=True, temperature=True, volume=True, speed=True))\n",
        "\n",
        "print(\"\\n> Simulating \" + str(nsteps) + \" steps...\")\n",
        "simulation.step(nsteps)\n",
        "\n",
        "simulation.reporters.clear() # remove all reporters so the next iteration don't trigger them.\n",
        "\n",
        "\n",
        "##################################\n",
        "# Writing last frame information of stride\n",
        "print(\"\\n> Writing state file (\" + str(rst_file) + \")...\")\n",
        "state = simulation.context.getState( getPositions=True, getVelocities=True )\n",
        "with open(rst_file, 'w') as f:\n",
        "\tf.write(XmlSerializer.serialize(state))\n",
        "\n",
        "last_frame = int(nsteps/nsavcrd)\n",
        "print(\"> Writing coordinate file (\" + str(pdb_file) + \", frame = \" + str(last_frame) + \")...\")\n",
        "positions = simulation.context.getState(getPositions=True).getPositions()\n",
        "PDBFile.writeFile(simulation.topology, positions, open(pdb_file, 'w'))\n",
        "\n",
        "print(\"\\n> Finished!\\n\")"
      ],
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "LXyL26HCO8Bu"
      },
      "source": [
        "---\n",
        "---\n",
        "# **Running a Production MD simulation**\n",
        "\n",
        "Finally, we will proceed with the Production simulation itself using the equilibrated system coordinates as input structure.\n",
        "\n",
        "Note that we will use here a *.rst state file* , which contains atomic velocities and positions from the last frame of the equilibration simulation, guaranteeing that our production simulation begins from a thermodynamically equilibrated system.\n",
        "\n",
        "Another important information here is the **Number_of_strides** and the **Stride_Time**. In this notebook, we simulate a defined number of *strides*, so the **simulation time = Number_of_strides*Stride_Time**. For example, we can simulate 100ns by setting *Number_of_strides=10* and *Stride_Time=10 ns*.\n",
        "\n",
        "**Important: at the end of the Production simulation, we concatenate all strides to create a complete trajectory file which can be visualized and analyzed**\n",
        "\n",
        "The idea behind this approach is to make use of the intermitent 12h/24h period that Google Colab allows us to use its GPUs."
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "Z0JV6Zid50_o",
        "cellView": "form"
      },
      "source": [
        "#@markdown ### **Provide input file names below:**\n",
        "\n",
        "Equilibrated_PDB = '1aki_equil.pdb' #@param {type:\"string\"}\n",
        "State_file = '1aki_equil.rst' #@param {type:\"string\"}\n",
        "\n",
        "#@markdown ---\n",
        "#@markdown ### **Parameters for MD Production protocol:**\n",
        "\n",
        "\n",
        "# remove whitespaces\n",
        "Jobname = '1aki_prod' #@param {type:\"string\"}\n",
        "\n",
        "#@markdown Simulation time (in nanoseconds), number of strides (integers) and integration timestep (in femtoseconds):\n",
        "Stride_Time = \"5\" #@param {type:\"string\"}\n",
        "stride_time_prod = Stride_Time\n",
        "Number_of_strides = \"1\" #@param {type:\"string\"}\n",
        "nstride = Number_of_strides\n",
        "Integration_timestep = \"2\" #@param [\"0.5\", \"1\", \"2\", \"3\", \"4\"]\n",
        "dt_prod = Integration_timestep\n",
        "\n",
        "#@markdown Temperature (in Kelvin) and Pressure (in bar)\n",
        "Temperature = 298 #@param {type:\"string\"}\n",
        "temperature_prod = Temperature\n",
        "Pressure = 1 #@param {type:\"string\"}\n",
        "pressure_prod = Pressure\n",
        "\n",
        "#@markdown Frequency to write the trajectory file (in picoseconds):\n",
        "Write_the_trajectory = \"10\" #@param [\"10\", \"100\", \"200\", \"500\", \"1000\"]\n",
        "write_the_trajectory_prod = Write_the_trajectory\n",
        "\n",
        "#@markdown Frequency to write the log file (in picoseconds):\n",
        "Write_the_log = \"10\" #@param [\"10\", \"100\", \"200\", \"500\", \"1000\"]\n",
        "write_the_log_prod = Write_the_log\n",
        "\n",
        "#@markdown ---"
      ],
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "0QcjKSoqHHhi",
        "cellView": "form"
      },
      "source": [
        "#@title **Runs a Production MD simulation (NPT ensemble) after equilibration**\n",
        "#\n",
        "###########################################\n",
        "import openmm as mm\n",
        "from openmm import *\n",
        "from openmm.app import *\n",
        "from openmm.unit import *\n",
        "\n",
        "from sys import stdout, exit, stderr\n",
        "import os, math, fnmatch\n",
        "\n",
        "#############################################\n",
        "# Defining MD simulation parameters\n",
        "\n",
        "jobname = os.path.join(workDir, str(Jobname))\n",
        "coordinatefile = os.path.join(workDir, str(GRO_filename))\n",
        "pdbfile = os.path.join(workDir, Equilibrated_PDB)\n",
        "topologyfile = os.path.join(workDir, str(TOP_filename))\n",
        "equil_rst_file = os.path.join(workDir, State_file)\n",
        "\n",
        "\n",
        "stride_time_ps = float(stride_time_prod)*1000\n",
        "stride_time = float(stride_time_ps)*picosecond\n",
        "nstride = int(Number_of_strides)\n",
        "dt = int(dt_prod)*femtosecond\n",
        "temperature = float(temperature_prod)*kelvin\n",
        "savcrd_freq = int(write_the_trajectory_prod)*picosecond\n",
        "print_freq  = int(write_the_log_prod)*picosecond\n",
        "\n",
        "pressure\t= float(pressure_prod)*bar\n",
        "\n",
        "simulation_time = stride_time*nstride\n",
        "nsteps  = int(stride_time.value_in_unit(picosecond)/dt.value_in_unit(picosecond))\n",
        "nprint  = int(print_freq.value_in_unit(picosecond)/dt.value_in_unit(picosecond))\n",
        "nsavcrd = int(savcrd_freq.value_in_unit(picosecond)/dt.value_in_unit(picosecond))\n",
        "firststride = 1 # must be integer\n",
        "#############################################\n",
        "# Defining functions to use below:\n",
        "def backup_old_log(pattern, string):\n",
        "\tresult = []\n",
        "\tfor root, dirs, files in os.walk(\"./\"):\n",
        "\t\tfor name in files:\n",
        "\t\t\tif fnmatch.fnmatch(name, pattern):\n",
        "\n",
        "\t\t\t\ttry:\n",
        "\t\t\t\t\tnumber = int(name[-2])\n",
        "\t\t\t\t\tavail = isinstance(number, int)\n",
        "\t\t\t\t\t#print(name,avail)\n",
        "\t\t\t\t\tif avail == True:\n",
        "\t\t\t\t\t\tresult.append(number)\n",
        "\t\t\t\texcept:\n",
        "\t\t\t\t\tpass\n",
        "\n",
        "\tif len(result) > 0:\n",
        "\t\tmaxnumber = max(result)\n",
        "\telse:\n",
        "\t\tmaxnumber = 0\n",
        "\n",
        "\tbackup_file = \"\\#\" + string + \".\" + str(maxnumber + 1) + \"#\"\n",
        "\tos.system(\"mv \" + string + \" \" + backup_file)\n",
        "\treturn backup_file\n",
        "##############################################\n",
        "\n",
        "#############################################\n",
        "print(\"\\n> Simulation details:\\n\")\n",
        "print(\"\\tJob name = \" + jobname)\n",
        "print(\"\\tCoordinate file = \" + str(coordinatefile))\n",
        "print(\"\\tPDB file = \" + str(pdbfile))\n",
        "print(\"\\tTopology file = \" + str(topologyfile))\n",
        "\n",
        "print(\"\\n\\tSimulation_time = \" + str(stride_time*nstride))\n",
        "print(\"\\tIntegration timestep = \" + str(dt))\n",
        "print(\"\\tTotal number of steps = \" +  str(nsteps*nstride))\n",
        "print(\"\\tNumber of strides = \" + str(nstride) + \" (\" + str(stride_time) + \" in each stride)\")\n",
        "\n",
        "print(\"\\n\\tSave coordinates each \" + str(savcrd_freq))\n",
        "print(\"\\tSave checkpoint each \" + str(savcrd_freq))\n",
        "print(\"\\tPrint in log file each \" + str(print_freq))\n",
        "\n",
        "print(\"\\n\\tTemperature = \" + str(temperature))\n",
        "print(\"\\tPressure = \" + str(pressure))\n",
        "#############################################\n",
        "\n",
        "print(\"\\n> Setting the system:\\n\")\n",
        "\n",
        "print(\"\\t- Reading topology and structure file...\")\n",
        "gro = GromacsGroFile(coordinatefile)\n",
        "top = GromacsTopFile(topologyfile, periodicBoxVectors=gro.getPeriodicBoxVectors(),\n",
        "        includeDir= os.path.join(workDir, str(foldername)))\n",
        "\n",
        "\n",
        "print(\"\\t- Creating system and setting parameters...\")\n",
        "nonbondedMethod = PME\n",
        "nonbondedCutoff = 1.0*nanometers\n",
        "ewaldErrorTolerance = 0.0005\n",
        "constraints = HBonds\n",
        "rigidWater = True\n",
        "constraintTolerance = 0.000001\n",
        "friction = 1.0\n",
        "system = top.createSystem(nonbondedMethod=nonbondedMethod, nonbondedCutoff=nonbondedCutoff,\n",
        "                           constraints=constraints, rigidWater=rigidWater, ewaldErrorTolerance=ewaldErrorTolerance)\n",
        "\n",
        "print(\"\\t- Setting barostat...\")\n",
        "system.addForce(MonteCarloBarostat(pressure, temperature))\n",
        "\n",
        "print(\"\\t- Setting integrator...\")\n",
        "integrator = LangevinIntegrator(temperature, friction, dt)\n",
        "integrator.setConstraintTolerance(constraintTolerance)\n",
        "simulation = Simulation(top.topology, system, integrator)\n",
        "simulation.context.setPositions(gro.positions)\n",
        "\n",
        "#############################################\n",
        "# Opening a loop of extension NSTRIDE to simulate the entire STRIDE_TIME*NSTRIDE\n",
        "for n in range(1, nstride + 1):\n",
        "\n",
        "\tprint(\"\\n\\n>>> Simulating Stride #\" + str(n) + \" <<<\")\n",
        "\n",
        "\tdcd_file = jobname + \"_\" + str(n) + \".dcd\"\n",
        "\tlog_file = jobname + \"_\" + str(n) + \".log\"\n",
        "\trst_file = jobname + \"_\" + str(n) + \".rst\"\n",
        "\tprv_rst_file = jobname + \"_\" + str(n-1) + \".rst\"\n",
        "\tpdb_file = jobname + \"_\" + str(n) + \".pdb\"\n",
        "\n",
        "\tif os.path.exists(rst_file):\n",
        "\t\tprint(\"> Stride #\" + str(n) + \" finished (\" + rst_file + \" present). Moving to next stride... <\")\n",
        "\t\tcontinue\n",
        "\n",
        "\tif n == 1:\n",
        "\t\tprint(\"\\n> Loading previous state from equilibration > \" + equil_rst_file + \" <\")\n",
        "\t\twith open(equil_rst_file, 'r') as f:\n",
        "\t\t\tsimulation.context.setState(XmlSerializer.deserialize(f.read()))\n",
        "\t\t\tcurrstep = int((n-1)*nsteps)\n",
        "\t\t\tcurrtime = currstep*dt.in_units_of(picosecond)\n",
        "\t\t\tsimulation.currentStep = currstep\n",
        "\t\t\tsimulation.context.setTime(currtime)\n",
        "\t\t\tprint(\"> Current time: \" + str(currtime) + \" (Step = \" + str(currstep) + \")\")\n",
        "\n",
        "\telse:\n",
        "\t\tprint(\"> Loading previous state from > \" + prv_rst_file + \" <\")\n",
        "\t\twith open(prv_rst_file, 'r') as f:\n",
        "\t\t\tsimulation.context.setState(XmlSerializer.deserialize(f.read()))\n",
        "\t\t\tcurrstep = int((n-1)*nsteps)\n",
        "\t\t\tcurrtime = currstep*dt.in_units_of(picosecond)\n",
        "\t\t\tsimulation.currentStep = currstep\n",
        "\t\t\tsimulation.context.setTime(currtime)\n",
        "\t\t\tprint(\"> Current time: \" + str(currtime) + \" (Step = \" + str(currstep) + \")\")\n",
        "\n",
        "\n",
        "\tdcd = DCDReporter(dcd_file, nsavcrd)\n",
        "\tfirstdcdstep = (currstep) + nsavcrd\n",
        "\tdcd._dcd = DCDFile(dcd._out, simulation.topology, simulation.integrator.getStepSize(), firstdcdstep, nsavcrd) # first step should not be 0\n",
        "\n",
        "\tsimulation.reporters.append(dcd)\n",
        "\tsimulation.reporters.append(StateDataReporter(stdout, nprint, step=True, speed=True, progress=True, totalSteps=(nsteps*nstride), remainingTime=True, separator='\\t\\t'))\n",
        "\tsimulation.reporters.append(StateDataReporter(log_file, nprint, step=True, kineticEnergy=True, potentialEnergy=True, totalEnergy=True, temperature=True, volume=True, speed=True))\n",
        "\n",
        "\tprint(\"\\n> Simulating \" + str(nsteps) + \" steps... (Stride #\" + str(n) + \")\")\n",
        "\tsimulation.step(nsteps)\n",
        "\n",
        "\tsimulation.reporters.clear() # remove all reporters so the next iteration don't trigger them.\n",
        "\n",
        "\n",
        "\t##################################\n",
        "\t# Writing last frame information of stride\n",
        "\tprint(\"\\n> Writing state file (\" + str(rst_file) + \")...\")\n",
        "\tstate = simulation.context.getState( getPositions=True, getVelocities=True )\n",
        "\twith open(rst_file, 'w') as f:\n",
        "\t\tf.write(XmlSerializer.serialize(state))\n",
        "\n",
        "\tlast_frame = int(nsteps/nsavcrd)\n",
        "\tprint(\"> Writing coordinate file (\" + str(pdb_file) + \", frame = \" + str(last_frame) + \")...\")\n",
        "\tpositions = simulation.context.getState(getPositions=True).getPositions()\n",
        "\tPDBFile.writeFile(simulation.topology, positions, open(pdb_file, 'w'))\n",
        "\n",
        "print(\"\\n> Finished!\\n\")"
      ],
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "DaLoQlJGf57o",
        "cellView": "form"
      },
      "source": [
        "#@title **Concatenate and align the trajectory**\n",
        "#@markdown **Important**: The **Google Drive Path**, **Jobname**, **Number of strides**, **stride time** and **trajectory saved frequency** should be the same you have been used to run your simulation in the previous steps.\n",
        "\n",
        "import MDAnalysis as mda\n",
        "from MDAnalysis.analysis import align, rms\n",
        "\n",
        "Google_Drive_Path = '/content/drive/MyDrive/GROMACS' #@param {type:\"string\"}\n",
        "workDir = Google_Drive_Path\n",
        "Equilibrated_PDB = '1aki_equil.pdb' #@param {type:\"string\"}\n",
        "Jobname = \"1aki_prod\" #@param {type: \"string\"}\n",
        "Skip = \"1\" #@param [\"1\", \"2\", \"5\", \"10\", \"20\", \"50\"]\n",
        "stride_traj = Skip\n",
        "Output_format = \"dcd\" #@param [\"dcd\", \"pdb\", \"trr\", \"xtc\"]\n",
        "first_stride = \"1\" #@param {type:\"string\"}\n",
        "Number_of_strides = \"1\" #@param {type:\"string\"}\n",
        "nstride = int(Number_of_strides)\n",
        "stride_time = \"5\" #@param {type:\"string\"}\n",
        "trajectory_saved_frequency = \"10\" #@param [\"10\", \"100\", \"200\", \"500\", \"1000\"]\n",
        "traj_save_freq = trajectory_saved_frequency\n",
        "Remove_waters = \"yes\" #@param [\"yes\", \"no\"]\n",
        "# stride_id_as_ref_for_alignment = \"1\" #@param {type: \"string\"}\n",
        "output_prefix = first_stride+\"-\"+str(int(first_stride)+nstride-1)\n",
        "\n",
        "stride_time_ps = float(stride_time)*1000\n",
        "simulation_time_analysis = stride_time_ps*nstride\n",
        "simulation_ns = float(stride_time)*int(Number_of_strides)\n",
        "number_frames = int(simulation_time_analysis)/int(traj_save_freq)\n",
        "number_frames_analysis = number_frames/int(stride_traj)\n",
        "\n",
        "\n",
        "# traj_end = os.path.join(workDir, str(Jobname) + \"_all.dcd\")\n",
        "nw_dcd = os.path.join(workDir, str(Jobname) + output_prefix + \"_nw.\" + str(Output_format))\n",
        "nw_pdb = os.path.join(workDir, str(Jobname) +  \"_nw.pdb\")\n",
        "whole_pdb = os.path.join(workDir, str(Jobname) +  \"_whole.pdb\")\n",
        "whole_dcd = os.path.join(workDir, str(Jobname) + output_prefix + \"_whole.\" + str(Output_format))\n",
        "template =  os.path.join(workDir, str(Jobname) + '_%s.dcd')\n",
        "pdb = os.path.join(workDir, Equilibrated_PDB)\n",
        "\n",
        "flist = [template % str(i) for i in range(int(first_stride), int(first_stride) + nstride)]\n",
        "ref = [template % int(1)]\n",
        "\n",
        "u1 = mda.Universe(pdb, flist)\n",
        "u2 = mda.Universe(pdb, ref)\n",
        "\n",
        "u2.trajectory[0] # set u2 to first frame\n",
        "\n",
        "# print(rms.rmsd(u1.select_atoms('name CA').positions, u2.select_atoms('name CA').positions, superposition=False))\n",
        "\n",
        "align.AlignTraj(u1, u2, select='name CA', in_memory=True).run()\n",
        "\n",
        "nw = u1.select_atoms(\"not (resname HOH)\")\n",
        "if Remove_waters == \"yes\":\n",
        "  with mda.Writer(nw_dcd, nw.n_atoms) as W:\n",
        "    for ts in u1.trajectory[::int(Skip)]:\n",
        "        W.write(nw, )\n",
        "  not_waters = u2.select_atoms(\"not (resname HOH)\")\n",
        "  not_waters.write(nw_pdb)\n",
        "  traj_dcd_check = os.path.exists(nw_dcd)\n",
        "  traj = nw_dcd\n",
        "  pdb_ref = nw_pdb\n",
        "else:\n",
        "  with mda.Writer(whole_dcd, u1.select_atoms(\"all\").n_atoms) as W:\n",
        "    for ts in u1.trajectory[::int(Skip)]:\n",
        "        W.write(u1.select_atoms(\"all\"))\n",
        "  whole = u2.select_atoms(\"all\")\n",
        "  whole.write(whole_pdb)\n",
        "  traj_dcd_check = os.path.exists(whole_dcd)\n",
        "  traj = whole_dcd\n",
        "  pdb_ref = whole_pdb\n",
        "\n",
        "traj_load = pt.load(traj, pdb_ref)\n",
        "print(traj_load)\n",
        "\n",
        "if traj_dcd_check == True:\n",
        "  print(\"Trajectory concatenated successfully! :-)\")\n",
        "else:\n",
        "  print(\"ERROR: Check your inputs! \")"
      ],
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "x_awbwdZnp6Q",
        "cellView": "form"
      },
      "source": [
        "#@title **Load, view and check the trajectory**\n",
        "#@markdown This will take a few minutes. Another coffee would be great. :-)\n",
        "\n",
        "#@markdown **Attention:** If your system has more than 100K atoms, you should skip this step. A high number of atoms can explode the memory on Colab.\n",
        "\n",
        "\n",
        "import warnings\n",
        "warnings.filterwarnings('ignore')\n",
        "\n",
        "#py3dmol functions\n",
        "class Atom(dict):\n",
        "  def __init__(self, line):\n",
        "    self[\"type\"] = line[0:6].strip()\n",
        "    self[\"idx\"] = line[6:11].strip()\n",
        "    self[\"name\"] = line[12:16].strip()\n",
        "    self[\"resname\"] = line[17:20].strip()\n",
        "    self[\"resid\"] = int(int(line[22:26]))\n",
        "    self[\"x\"] = float(line[30:38])\n",
        "    self[\"y\"] = float(line[38:46])\n",
        "    self[\"z\"] = float(line[46:54])\n",
        "    self[\"sym\"] = line[76:78].strip()\n",
        "\n",
        "  def __str__(self):\n",
        "    line = list(\" \" * 80)\n",
        "    line[0:6] = self[\"type\"].ljust(6)\n",
        "    line[6:11] = self[\"idx\"].ljust(5)\n",
        "    line[12:16] = self[\"name\"].ljust(4)\n",
        "    line[17:20] = self[\"resname\"].ljust(3)\n",
        "    line[22:26] = str(self[\"resid\"]).ljust(4)\n",
        "    line[30:38] = str(self[\"x\"]).rjust(8)\n",
        "    line[38:46] = str(self[\"y\"]).rjust(8)\n",
        "    line[46:54] = str(self[\"z\"]).rjust(8)\n",
        "    line[76:78] = self[\"sym\"].rjust(2)\n",
        "    return \"\".join(line) + \"\\n\"\n",
        "\n",
        "class Molecule(list):\n",
        "  def __init__(self, file):\n",
        "    for line in file:\n",
        "      if \"ATOM\" in line or \"HETATM\" in line:\n",
        "        self.append(Atom(line))\n",
        "\n",
        "    def __str__(self):\n",
        "      outstr = \"\"\n",
        "      for at in self:\n",
        "        outstr += str(at)\n",
        "      return outstr\n",
        "\n",
        "if number_frames_analysis > 10:\n",
        "  stride_animation = number_frames_analysis/10\n",
        "else:\n",
        "  stride_animation = 1\n",
        "\n",
        "u = mda.Universe(pdb_ref, traj)\n",
        "\n",
        "# Write out frames for animation\n",
        "protein = u.select_atoms('not (resname SOL)')\n",
        "i = 0\n",
        "for ts in u.trajectory[0:len(u.trajectory):int(stride_animation)]:\n",
        "    if i > -1:\n",
        "        with mda.Writer('' + str(i) + '.pdb', protein.n_atoms) as W:\n",
        "            W.write(protein)\n",
        "    i = i + 1\n",
        "# Load frames as molecules\n",
        "molecules = []\n",
        "for i in range(int(len(u.trajectory)/int(stride_animation))):\n",
        "    with open('' + str(i) + '.pdb') as ifile:\n",
        "        molecules.append(Molecule(ifile))\n",
        "\n",
        "models = \"\"\n",
        "for i in range(len(molecules)):\n",
        "  models += \"MODEL \" + str(i) + \"\\n\"\n",
        "  for j,mol in enumerate(molecules[i]):\n",
        "    models += str(mol)\n",
        "  models += \"ENDMDL\\n\"\n",
        "#view.addModelsAsFrames(models)\n",
        "\n",
        "# Animation\n",
        "view = py3Dmol.view(width=800, height=600)\n",
        "view.addModelsAsFrames(models)\n",
        "for i, at in enumerate(molecules[0]):\n",
        "    default = {\"cartoon\": {'color': 'spectrum'}}\n",
        "    view.setStyle({'model': -1, 'serial': i+1}, at.get(\"pymol\", default))\n",
        "\n",
        "view.zoomTo()\n",
        "view.animate({'loop': \"forward\"})\n",
        "view.show()"
      ],
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "Emh0vU5UjgB6"
      },
      "source": [
        "---\n",
        "---\n",
        "# **Analysis**\n",
        "\n",
        "Although visualizing your trajectory can be quite useful, sometimes you also want more quantitative data.\n",
        "\n",
        "Analyses of MD trajectories vary a lot and we do not intend to cover it all here. However, one can make use of MDanalysis or PyTraj to easily analyze simulations.\n",
        "\n",
        "Below, you can find a few examples of code snippets that can help you to shed some light on your simulation behavior."
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "wBrBMF4Puyv6",
        "cellView": "form"
      },
      "source": [
        "#@title **Compute RMSD of protein's CA atoms**\n",
        "#@markdown **Provide output file names below:**\n",
        "Output_name = 'rmsd_ca' #@param {type:\"string\"}\n",
        "\n",
        "\n",
        "rmsd = pt.rmsd(traj_load, ref = 0, mask = \"@CA\")\n",
        "\n",
        "time = len(rmsd)*int(Write_the_trajectory)/1000\n",
        "time_array = np.arange(0,time,int(Write_the_trajectory)/1000)*int(stride_traj)\n",
        "\n",
        "# Plotting:\n",
        "ax = plt.plot(time_array, rmsd, alpha=0.6, color = 'blue', linewidth = 1.0)\n",
        "plt.xlim(0, simulation_ns)\n",
        "#plt.ylim(2, 6)\n",
        "\n",
        "plt.xlabel(\"Time (ns)\", fontsize = 14, fontweight = 'bold')\n",
        "plt.ylabel(\"RMSD [$\\AA$]\", fontsize = 14, fontweight = 'bold')\n",
        "plt.xticks(fontsize = 12)\n",
        "plt.yticks(fontsize = 12)\n",
        "plt.savefig(os.path.join(workDir, Output_name + \".png\"), dpi=600, bbox_inches='tight')\n",
        "\n",
        "raw_data=pd.DataFrame(rmsd)\n",
        "raw_data.to_csv(os.path.join(workDir, Output_name + \".csv\"))"
      ],
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "ZHyMpikjuaLT",
        "cellView": "form"
      },
      "source": [
        "#@title **Plot RMSD as a ditribution**\n",
        "\n",
        "#@markdown **Provide output file names below:**\n",
        "Output_name = 'rmsd_dist' #@param {type:\"string\"}\n",
        "\n",
        "ax = sb.kdeplot(rmsd, color=\"blue\", shade=True, alpha=0.2, linewidth=0.5)\n",
        "plt.xlabel('RMSD [$\\AA$]', fontsize = 14, fontweight = 'bold')\n",
        "plt.xticks(fontsize = 12)\n",
        "plt.yticks([])\n",
        "plt.ylabel('')\n",
        "ax.spines['top'].set_visible(False)\n",
        "ax.spines['right'].set_visible(False)\n",
        "ax.spines['bottom'].set_visible(True)\n",
        "ax.spines['left'].set_visible(False)\n",
        "\n",
        "plt.savefig(os.path.join(workDir, Output_name + \".png\"), dpi=600, bbox_inches='tight')"
      ],
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "CvOFrXGXwXrV",
        "cellView": "form"
      },
      "source": [
        "#@title **Compute radius of gyration of protein's CA atoms**\n",
        "\n",
        "#@markdown **Provide output file names below:**\n",
        "Output_name = 'radius_gyration' #@param {type:\"string\"}\n",
        "\n",
        "radgyr = pt.radgyr(traj_load, mask = \"@CA\")\n",
        "time = len(rmsd)*int(Write_the_trajectory)/1000\n",
        "time_array = np.arange(0,time,int(Write_the_trajectory)/1000)*int(stride_traj)\n",
        "\n",
        "# Plotting:\n",
        "plt.plot(time_array, radgyr, alpha=0.6, color = 'green', linewidth = 1.0)\n",
        "plt.xlim(0, simulation_ns)\n",
        "#plt.ylim(2, 6)\n",
        "\n",
        "plt.xlabel(\"Time (ns)\", fontsize = 14, fontweight = 'bold')\n",
        "plt.ylabel(\"Radius of gyration ($\\AA$)\", fontsize = 14, fontweight = 'bold')\n",
        "plt.xticks(fontsize = 12)\n",
        "plt.yticks(fontsize = 12)\n",
        "plt.savefig(os.path.join(workDir, Output_name + \".png\"), dpi=600, bbox_inches='tight')\n",
        "\n",
        "raw_data=pd.DataFrame(radgyr)\n",
        "raw_data.to_csv(os.path.join(workDir, Output_name + \".csv\"))"
      ],
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "2Q7FKg8Fuxr9",
        "cellView": "form"
      },
      "source": [
        "#@title **Plot radius of gyration as a ditribution**\n",
        "\n",
        "#@markdown **Provide output file names below:**\n",
        "Output_name = 'radius_gyration_dist' #@param {type:\"string\"}\n",
        "\n",
        "ax = sb.kdeplot(radgyr, color=\"green\", shade=True, alpha=0.2, linewidth=0.5)\n",
        "plt.xlabel('Radius of gyration ($\\AA$)', fontsize = 14, fontweight = 'bold')\n",
        "plt.xticks(fontsize = 12)\n",
        "plt.yticks([])\n",
        "plt.ylabel('')\n",
        "ax.spines['top'].set_visible(False)\n",
        "ax.spines['right'].set_visible(False)\n",
        "ax.spines['bottom'].set_visible(True)\n",
        "ax.spines['left'].set_visible(False)\n",
        "\n",
        "plt.savefig(os.path.join(workDir, Output_name + \".png\"), dpi=600, bbox_inches='tight')"
      ],
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "p2Y0DgwTxLWc",
        "cellView": "form"
      },
      "source": [
        "#@title **Compute RMSF of protein's CA atoms**\n",
        "\n",
        "#@markdown **Provide output file names below:**\n",
        "Output_name = 'rmsf_ca' #@param {type:\"string\"}\n",
        "\n",
        "\n",
        "rmsf = pt.rmsf(traj_load, \"@CA\")\n",
        "bfactor = pt.bfactors(traj_load, byres=True)\n",
        "\n",
        "# Plotting:\n",
        "plt.plot(rmsf[:,1], alpha=1.0, color = 'red', linewidth = 1.0)\n",
        "\n",
        "plt.xlabel(\"Residue\", fontsize = 14, fontweight = 'bold')\n",
        "plt.ylabel(\"RMSF ($\\AA$)\", fontsize = 14, fontweight = 'bold')\n",
        "plt.xticks(fontsize = 12)\n",
        "plt.xlim(0, len(rmsf[:-1]))\n",
        "\n",
        "#plt.xticks(np.arange(min(rmsf[:1]), max(rmsf[:1])))\n",
        "plt.yticks(fontsize = 12)\n",
        "plt.savefig(os.path.join(workDir, Output_name + \".png\"), dpi=600, bbox_inches='tight')\n",
        "\n",
        "raw_data=pd.DataFrame(rmsf)\n",
        "raw_data.to_csv(os.path.join(workDir, Output_name + \".csv\"))"
      ],
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "JalicqqrTodW",
        "cellView": "form"
      },
      "source": [
        "#@title **2D RMSD**\n",
        "\n",
        "#@markdown **Provide output file names below:**\n",
        "Output_name = '2D_rmsd' #@param {type:\"string\"}\n",
        "\n",
        "last_frame = len(time_array)\n",
        "\n",
        "stride_ticks_f = (last_frame)/5\n",
        "ticks_frame = np.arange(0,(len(time_array) + float(stride_ticks_f)), float(stride_ticks_f))\n",
        "a = ticks_frame.astype(float)\n",
        "stride_ticks_t = (simulation_ns)/5\n",
        "tick_time = np.arange(0,(float(simulation_ns) + float(stride_ticks_t)), float(stride_ticks_t))\n",
        "b = tick_time.astype(float)\n",
        "\n",
        "mat1 = pt.pairwise_rmsd(traj_load, mask=\"@CA\", frame_indices=range(int(number_frames_analysis)))\n",
        "\n",
        "\n",
        "ax = plt.imshow(mat1, cmap = 'PRGn', origin='lower', interpolation = 'bicubic')\n",
        "plt.title('2D RMSD')\n",
        "plt.xlabel('Time (ns)', fontsize = 14, fontweight = 'bold')\n",
        "plt.ylabel('Time (ns)', fontsize = 14, fontweight = 'bold')\n",
        "# plt.xticks(fontsize = 12)\n",
        "# plt.yticks(fontsize = 12)\n",
        "plt.xticks(a, b.round(decimals=3), fontsize = 12)\n",
        "plt.yticks(a, b.round(decimals=3), fontsize = 12)\n",
        "# plt.xlim(0, a[-1])\n",
        "# plt.ylim(0, a[-1])\n",
        "\n",
        "cbar1 = plt.colorbar()\n",
        "cbar1.set_label(\"RMSD ($\\AA$)\", fontsize = 14, fontweight = 'bold')\n",
        "\n",
        "\n",
        "plt.savefig(os.path.join(workDir, Output_name + \".png\"), dpi=600, bbox_inches='tight')\n",
        "\n",
        "raw_data=pd.DataFrame(mat1)\n",
        "raw_data.to_csv(os.path.join(workDir, Output_name + \".csv\"))"
      ],
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "_mgVSbBshWFV",
        "cellView": "form"
      },
      "source": [
        "#@title **Calculate eigvenctors of Principle Component Analysis (PCA)**\n",
        "data = pt.pca(traj_load, fit=True, ref=0, mask='@CA', n_vecs=2)\n",
        "#print('projection values of each frame to first mode = {} \\n'.format(data[0][0]))\n",
        "#print('projection values of each frame to second mode = {} \\n'.format(data[0][1]))\n",
        "#print('eigvenvalues of first two modes', data[1][0])\n",
        "#print(\"\")\n",
        "#print('eigvenvectors of first two modes: \\n', data[1][1])\n",
        "\n",
        "last_frame = len(time_array)\n",
        "\n",
        "stride_ticks_f = (last_frame)/5\n",
        "ticks_frame = np.arange(0,(len(time_array) + float(stride_ticks_f)), float(stride_ticks_f))\n",
        "a = ticks_frame.astype(float)\n",
        "a2 = a.tolist()\n",
        "stride_ticks_t = (simulation_ns)/5\n",
        "tick_time = np.arange(0,(float(simulation_ns) + float(stride_ticks_t)), float(stride_ticks_t))\n",
        "b = tick_time.astype(float)\n",
        "\n",
        "#@markdown **Provide output file names below:**\n",
        "Output_name = 'PCA' #@param {type:\"string\"}\n",
        "\n",
        "Output_PC1 = 'PC1' #@param {type:\"string\"}\n",
        "Output_PC2 = 'PC2' #@param {type:\"string\"}\n",
        "\n",
        "%matplotlib inline\n",
        "%config InlineBackend.figure_format = 'retina'  # high resolution\n",
        "projection_data = data[0]\n",
        "plt.title(r'PCA of C-$\\alpha$')\n",
        "PC1 = data[0][0]\n",
        "PC2 = data[0][1]\n",
        "\n",
        "a = plt.scatter(PC1,PC2, c=range(int(number_frames_analysis)), cmap='Greens', marker='o',s=8, alpha=1)\n",
        "plt.clim(0, last_frame)\n",
        "\n",
        "plt.xlabel('PC1', fontsize = 14, fontweight = 'bold')\n",
        "plt.ylabel('PC2', fontsize = 14, fontweight = 'bold')\n",
        "plt.xticks(fontsize = 12)\n",
        "plt.yticks(fontsize = 12)\n",
        "# N = len(number_frames)\n",
        "# x2 = np.arange(N)\n",
        "\n",
        "cbar1 = plt.colorbar(a, orientation=\"vertical\")\n",
        "cbar1.set_label('Time(ns)', fontsize = 14, fontweight = 'bold')\n",
        "cbar1.set_ticks(a2)\n",
        "cbar1.set_ticklabels(b.round(decimals=3))\n",
        "\n",
        "plt.savefig(os.path.join(workDir, Output_name + \".png\"), dpi=600, bbox_inches='tight')\n",
        "\n",
        "pc1=pd.DataFrame(PC1)\n",
        "pc1.to_csv(os.path.join(workDir, Output_PC1 + \".csv\"))\n",
        "pc2=pd.DataFrame(PC2)\n",
        "pc2.to_csv(os.path.join(workDir, Output_PC2 + \".csv\"))"
      ],
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "yce9RfNtpl-J",
        "cellView": "form"
      },
      "source": [
        "#@title **Plot Principal Component 1 (PC1) and Principal Component 2 (PC2) as a ditribution**\n",
        "Output_name = 'PCA_dist' #@param {type:\"string\"}\n",
        "\n",
        "\n",
        "fig = plt.figure(figsize=(9,5))\n",
        "\n",
        "plt.subplot(1, 2, 1)\n",
        "ax = sb.kdeplot(PC1, color=\"green\", shade=True, alpha=0.2, linewidth=0.5)\n",
        "plt.xlabel('PC1', fontsize = 14, fontweight = 'bold')\n",
        "plt.xticks(fontsize = 12)\n",
        "plt.yticks([])\n",
        "plt.ylabel('')\n",
        "ax.spines['top'].set_visible(False)\n",
        "ax.spines['right'].set_visible(False)\n",
        "ax.spines['bottom'].set_visible(True)\n",
        "ax.spines['left'].set_visible(False)\n",
        "plt.subplot(1, 2, 2)\n",
        "ax2 = sb.kdeplot(PC2, color=\"purple\", shade=True, alpha=0.2, linewidth=0.5)\n",
        "plt.xlabel('PC2', fontsize = 14, fontweight = 'bold')\n",
        "plt.xticks(fontsize = 12)\n",
        "plt.yticks([])\n",
        "plt.ylabel('')\n",
        "ax2.spines['top'].set_visible(False)\n",
        "ax2.spines['right'].set_visible(False)\n",
        "ax2.spines['bottom'].set_visible(True)\n",
        "ax2.spines['left'].set_visible(False)\n",
        "\n",
        "\n",
        "plt.savefig(os.path.join(workDir, Output_name + \".png\"), dpi=600, bbox_inches='tight')"
      ],
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "pTDb7CEfkLq1",
        "cellView": "form"
      },
      "source": [
        "#@title **Pearson's Cross Correlation (CC)**\n",
        "\n",
        "#@markdown **Provide output file names below:**\n",
        "Output_name = 'cross_correlation' #@param {type:\"string\"}\n",
        "\n",
        "\n",
        "traj_align = pt.align(traj_load, mask='@CA', ref=0)\n",
        "\n",
        "mat_cc = matrix.correl(traj_align, '@CA')\n",
        "\n",
        "ax = plt.imshow(mat_cc, cmap = 'PiYG_r', interpolation = 'bicubic', vmin = -1, vmax = 1, origin='lower')\n",
        "\n",
        "plt.xlabel('Residues', fontsize = 14, fontweight = 'bold')\n",
        "plt.ylabel('Residues', fontsize = 14, fontweight = 'bold')\n",
        "plt.xticks(fontsize = 12)\n",
        "plt.yticks(fontsize = 12)\n",
        "cbar1 = plt.colorbar()\n",
        "cbar1.set_label('$CC_ij$', fontsize = 14, fontweight = 'bold')\n",
        "\n",
        "plt.savefig(os.path.join(workDir, Output_name + \".png\"), dpi=600, bbox_inches='tight')\n",
        "\n",
        "raw_data=pd.DataFrame(mat_cc)\n",
        "raw_data.to_csv(os.path.join(workDir, Output_name + \".csv\"))"
      ],
      "execution_count": null,
      "outputs": []
    }
  ]
}