2018/Biological_Computing_with_Synthetic_Minimal_Cells_and_Quorum_Sensing,_Sci_fair_2021.ipynb

{
  "nbformat": 4,
  "nbformat_minor": 0,
  "metadata": {
    "colab": {
      "name": "Biological Computing with Synthetic Minimal Cells and Quorum Sensing, Sci fair 2021",
      "provenance": [],
      "collapsed_sections": []
    },
    "kernelspec": {
      "name": "python3",
      "display_name": "Python 3"
    }
  },
  "cells": [
    {
      "cell_type": "code",
      "metadata": {
        "id": "LbVaq0saa5sA"
      },
      "source": [
        "'''\r\n",
        "My science fair project! Take a look at https://github.com/dhrumilp15/Science-Fair/ for more docs!\r\n",
        "Author: Dhrumil Patel\r\n",
        "'''\r\n",
        "#import libraries\r\n",
        "import numpy as np\r\n",
        "from scipy.integrate import odeint\r\n",
        "import matplotlib.pyplot as plt\r\n",
        "%matplotlib inline\r\n",
        "from itertools import chain"
      ],
      "execution_count": 1,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "bbC5N5bXbBW7"
      },
      "source": [
        "# ODE\r\n",
        "alpha_A = 1.0564*(10**-7) #  AHL production rate\r\n",
        "alpha = 1.0564 * (10**-7)\r\n",
        "alpha_R = 1.0 * (10**-7)# Production of PpuR\r\n",
        "\r\n",
        "beta_A = 1.0564*(10**-6) # Feedback-regulated AHL production rate\r\n",
        "beta = 1.0564*(10**-6)\r\n",
        "\r\n",
        "x_0 = 5.4044 * (10**-7) # Initial value of AHL Concentration\r\n",
        "R_0 = 1.0 # Initial value of PpuR Concentration\r\n",
        "y_0 = 5.2 * (10**3) #Initial amount of Lactonase\r\n",
        "n_0 = 1.0 # Size of biomass\r\n",
        "s_0 = 1.0 # Amount of substrate\r\n",
        "\r\n",
        "n = 2.3 # Hill coefficient for x\r\n",
        "n_s = 2.5 # Hill Coefficient for substrate consumption process\r\n",
        "\r\n",
        "m = 2.5 # Hill coefficient\r\n",
        "\r\n",
        "gamma_A = 1.0 #Threshold of alpha_A\r\n",
        "gamma_R = 1.0 #Threshold of alpha_R\r\n",
        "gamma_S = 1.0 # Decay rate of substrate - Rate of substrate consumption\r\n",
        "gamma = 0.105 #AHL Decay rate\r\n",
        "\r\n",
        "pi_1_plus = 1e-7 # Rate at which AHL binds to PpuR\r\n",
        "pi_minus = 10.0 # Rate at which AHL and PpuR split up(disassociate)\r\n",
        "\r\n",
        "C_0 = 1.0 # Initial concentration of PpuR-AHL complex\r\n",
        "C_th = 3.597 * (10**-13) # Threshold of positive feedback for AHL concentration, x_th\r\n",
        "C_th2 = 1.0 #Threshold for lactonase activation\r\n",
        "\r\n",
        "D = 2.5 # Dilution Rate\r\n",
        "\r\n",
        "S = 1.0 # Concentration of Substrate\r\n",
        "\r\n",
        "K_m = 1.0 # Carrying capacity, consumption rate is reached\r\n",
        "\r\n",
        "tau = 2 # Delay in production of lactonase\r\n",
        "delta = 1.5*(10**-4) # Lactonase-dependent AHL degradation rate\r\n",
        "omega = 0.105 # Lactonase decay rate(w/ washout)\r\n",
        "rho = 5.0521 * (10**3) # Lactonase Production Rate\r\n",
        "\r\n",
        "N = 4.5929 * (10**11) #Cell density at equilibrium\r\n",
        "\r\n",
        "x_th = 3.597 * (10**-12) #Critical threshold for positive-feedback in x\r\n",
        "y_th = 3.597 * (10**-12) #Critical threshold for positive-feedback in y\r\n",
        "\r\n",
        "#spanish_gamma = (pi_1_plus*alpha_R)/(pi_minus*gamma_R)\r\n",
        "\r\n",
        "p = [alpha_A, beta_A, n, gamma_A, gamma_R, pi_1_plus, pi_minus, C_th, alpha_R]\r\n",
        "\r\n",
        "# ODE solver parameters\r\n",
        "abserr = 1.0e-8\r\n",
        "relerr = 1.0e-6\r\n",
        "stoptime = 10.0\r\n",
        "numpoints = 2000\r\n",
        "\r\n",
        "t = [stoptime * float(i) / (numpoints - 1) for i in range(numpoints)] # How long the graph will be\r\n",
        "\r\n",
        "def single_cell_model(w0,t,p, q): # Initial Values of Functions user gives\r\n",
        "  x_0, R_0, C_0 = w0 # We redefine the \"initial values\" of the function on each iteration in odeint\r\n",
        "  \r\n",
        "  alpha_A, beta_A, n, gamma_A, gamma_R, pi_1_plus, pi_minus, C_th, alpha_R = p # Parameters\r\n",
        "  q = q # So that q can be introduced in AHL Concentration function\r\n",
        "  f = [alpha_A + beta_A*((C_0**n)/(C_th**n + C_0**n))-gamma_A*x_0 - pi_1_plus*x_0*R_0 + pi_minus*C_0 - q*x_0,\r\n",
        "       alpha_R - pi_1_plus* x_0* R_0 + pi_minus*C_0 - gamma_R*R_0,\r\n",
        "       pi_1_plus*x_0*R_0 - pi_minus*C_0] # 1st Line - AHL Concentration, 2nd Line - PpuR Concentration, 3rd Line - PpuR-AHL Complex Concentration\r\n",
        "  return f # Put functions in f to be integrated in solver()\r\n",
        "\r\n",
        "  \"\"\"\r\n",
        "  1st line - Substrate concentration: D*s_0 - gamma_S * n_0 ((s_0**n_s)/((K_m ** n_s) + s_0 ** n_s)) - D*S\r\n",
        "  \r\n",
        "  2nd line - Biomass size: alpha_A * n_0 * ( (s_0 ** n_s) / ((K_m ** n_s) + s_0 ** n_s)) - D*n_0\r\n",
        "  \r\n",
        "  3rd  line - AHL Concentration: (alpha_A + beta_A*((C_0**n)/((C_th**n) + (C_0**n))))*n_0 - gamma_A*x_0 - pi_1_plus*R_0*x_0 + pi_1_minus*C_0 - D*x_0 - delta*x_0*y_0\r\n",
        "  \r\n",
        "  4th line - PpuR-AHL Concentration: pi_1_plus*R_0*x_0 - pi_1_minus*C_0\r\n",
        "  \r\n",
        "  5th line - PpuR Concentration: alpha_R + pi_1_minus*C_0 - pi_1_plus*R_0*x_0 - gamma_R*R_0\r\n",
        "  \r\n",
        "  Initial values of functions(will go to w1): s_0, n_0, C_0, R_0, x_0, y_0\r\n",
        "  \r\n",
        "  Variables(will go to p1): D, gamma_S, S, n_s, K_m, alpha_A, N, beta_A, n, C_th, gamma_A, pi_1_plus, pi_1_minus, delta, alpha_R, gamma_R\r\n",
        "    \r\n",
        "  Lactonase concentration: alpha_L*(C(t - tau)**m)/(C_th2**m + C(t - tau)**m)*n_0 - gamma_L * y_0 - D * y_0\r\n",
        "  \"\"\"\r\n",
        "\r\n",
        "# def single_cell_full_model(w1, t, p1, q):\r\n",
        "#   w1 = C_0, R_0, x_0, y_0, n_0, s_0\r\n",
        "#   print(p1)\r\n",
        "  \r\n",
        "#   D, gamma_S, S, n_s, m, K_m, alpha_A, beta_A, n, C_th, C_th2, gamma_A, pi_1_plus, pi_minus, delta, alpha_R, gamma_R = p1\r\n",
        "  \r\n",
        "#   fm = [D * S - gamma_S * n_0*((s_0 ** n_s)/((K_m ** n_s) + (s_0 ** n_s))) - D * s_0,\r\n",
        "#         alpha_A * n_0 * ( (s_0 ** n_s) / ((K_m ** n_s) + (s_0 ** n_s))) - D * n_0,\r\n",
        "#         (alpha_A + beta_A * ((C_0 ** n)/((C_th ** n) + (C_0 ** n)))) * n_0 - gamma_A * x_0 - pi_1_plus * R_0 * x_0 + pi_minus * C_0 - D * x_0 - delta * x_0 * y_0,\r\n",
        "#         pi_1_plus * R_0 * x_0 - pi_minus * C_0,\r\n",
        "#         alpha_R + pi_minus * C_0 - pi_1_plus * R_0 * x_0 - gamma_R * R_0,\r\n",
        "#        alpha_L*(C_0**m)/(C_th2**m + C_0**m)*n_0 - gamma_L * y_0 - D * y_0]\r\n",
        "#   return fm\r\n",
        "\r\n",
        "class Bacterium:\r\n",
        "  def __init__(self,x_0, R_0, C_0):\r\n",
        "    self.C_0 = C_0 # PpuR-AHL\r\n",
        "    self.x_0 = x_0 # AHL concentration\r\n",
        "    self.R_0 = R_0 # PpuR Concentration\r\n",
        "    #self.y_0 = y_0 # Lactonase concentration\r\n",
        "    #self.n_0 = n_0 # Size of biomass\r\n",
        "    #self.s_0 = s_0 # Substrate\r\n",
        "    self.w0 = [self.x_0, self.R_0, self.C_0]\r\n",
        "    #self.w1 = [self.x_0, self.R_0, self.C_0, self.n_0, self.s_0, self.y_0]\r\n",
        "  \r\n",
        "  def solver(self):\r\n",
        "    wsol = []\r\n",
        "    for q in np.linspace(0.0, 3.0, num=100):\r\n",
        "      wsol.append(odeint(single_cell_model, self.w0, t, args=(p, q), atol = abserr, rtol = relerr))\r\n",
        "      # Atol and rtol are tolerants\r\n",
        "    return wsol\r\n",
        "  \r\n",
        "  #def solver1(self):\r\n",
        "    #wsol1 = odeint(full_model, self.w1, t, args=(p1,), atol = abserr, rtol = relerr)\r\n",
        "    #return wsol1\r\n",
        "    \r\n",
        "  def nan_helper(self, A): # Takes care of NaNs\r\n",
        "      ok = ~np.isnan(A) #Are there NaNs\r\n",
        "      xp = ok.ravel().nonzero()[0] # x-coord of data values ravel for scaling up larger w/ multi-dim data\r\n",
        "      fp = A[~np.isnan(A)]# y-coord of data values\r\n",
        "      x  = np.isnan(A).ravel().nonzero()[0] # x-coord of interp values\r\n",
        "      A[np.isnan(A)] = np.interp(x, xp, fp)\r\n",
        "      return A\r\n",
        "  \r\n",
        "  def ranger(self,x, axis=1):\r\n",
        "    return np.max(x, axis=axis) - np.min(x, axis=axis)\r\n",
        "  \r\n",
        "  def plot(self): #Function to plot graph\r\n",
        "    \r\n",
        "    # Two subplots, the axes array is 1-d\r\n",
        "    N = 2 # Used only in this method\r\n",
        "    #f, axarr = plt.subplots(N, sharex=False)\r\n",
        "    x_new = []\r\n",
        "    for i in range(0,100):\r\n",
        "      x_new.append(self.solver()[i][:,0])\r\n",
        "    \r\n",
        "    ### Finding Double Difference between original AHl concentration and ones with AHL degrading enzyme\r\n",
        "    x_diff = []\r\n",
        "    x_new[0] = self.nan_helper(np.asarray(x_new[0]))\r\n",
        "    for i in range(1,100):\r\n",
        "      x_new[i] = self.nan_helper(np.asarray(x_new[i]))\r\n",
        "      x_diff.append(np.gradient(x_new[0] - x_new[i])) \r\n",
        "    x_range = []\r\n",
        "    \r\n",
        "    #find ranges of values - ideally, we want something with a small range\r\n",
        "    for i in range(len(x_diff)):\r\n",
        "      x_range.append(np.ptp(x_diff[i]))\r\n",
        "    \r\n",
        "    print (np.where(np.square(x_diff[1]) < 0.0025))\r\n",
        "    \r\n",
        "    #for i in range(N):\r\n",
        "      #axarr[i].plot(t, x_new[50*i], 'b')\r\n",
        "      #axarr[i].set_title('AHL Concentration')\r\n",
        "      #axarr[i].tick_params(\r\n",
        "      #axis='x',          # changes apply to the x-axis\r\n",
        "      #which='both',      # both major and minor ticks are affected\r\n",
        "      #bottom='off',      # ticks along the bottom edge are off\r\n",
        "      #labelbottom='off')\r\n",
        "    \r\n",
        "    # What the graph is going to look like\r\n",
        "    original = plt.plot(t,x_new[0],'g', label = 'Original') # Graph with q set to 0\r\n",
        "    q06 = plt.plot(t,x_new[20],'c', label = 'Degradation Rate = 0.61') # Graph with q set to 0.61\r\n",
        "    q15 = plt.plot(t,x_new[50],'b', label = 'Degradation Rate = 1.5') # Graph with q set to 1.5 - EXTRANEOUS; AHL hits 0 :(\r\n",
        "    q1 = plt.plot(t,x_new[99],'r', label = 'Degradation Rate = 3.0')# Graph with q set to 3  - EXTRANEOUS; AHL hits 0 :( \r\n",
        "    plt.legend(loc = 'upper right') # Put graph in upper right corner\r\n",
        "    plt.xlabel('Time (h)') # Label of x axis\r\n",
        "    plt.ylabel('AHL Concentration (M)') # Label of y axis\r\n",
        "    plt.title('Performance of Tunable Biological Resistor') # Title\r\n",
        "    plt.show() # Show graph"
      ],
      "execution_count": 2,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 329
        },
        "id": "F1lxodsNfOZS",
        "outputId": "afb696f0-430e-4725-8e84-7e7aceebc0a8"
      },
      "source": [
        "bob = Bacterium(x_0, R_0, C_0) # Initial Values assigned at beg. of code\r\n",
        "bob.plot() # Time to plot"
      ],
      "execution_count": 3,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "/usr/local/lib/python3.7/dist-packages/ipykernel_launcher.py:64: RuntimeWarning: invalid value encountered in double_scalars\n"
          ],
          "name": "stderr"
        },
        {
          "output_type": "stream",
          "text": [
            "(array([   0,    1,    2, ..., 1997, 1998, 1999]),)\n"
          ],
          "name": "stdout"
        },
        {
          "output_type": "display_data",
          "data": {
            "image/png": "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\n",
            "text/plain": [
              "<Figure size 432x288 with 1 Axes>"
            ]
          },
          "metadata": {
            "tags": [],
            "needs_background": "light"
          }
        }
      ]
    }
  ]
}