{
 "cells": [
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# ageas.Plot()\n",
    "\n",
    "This notebook demonstrate how to use ageas.Plot() to visualize GRNs or regulons reconstructed with key GRPs extracted with AGEAS."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "import ageas"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "If visualization will be done in script after ageas.Launch() or ageas.Unit(), initialization Plot() object could be as simple as:\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "# ageas.Test() is a Launch() object using internalized data\n",
    "# This would take a few minutes to run\n",
    "test_launch_object = ageas.Test(cpu_mode = True)\n",
    "\n",
    "# Here we use first network in atlas, network_0 for example\n",
    "network = test_launch_object.atlas.nets[0]"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Otherwise, we can load in networks from ***full_atlas.js*** or ***key_atalas.js*** output by [ageas.Launch()](https://JackSSK.github.io/Ageas/html/generated/ageas.Launch.html)."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "import ageas.tool.grn as grn\n",
    "import ageas.tool.json as json\n",
    "\n",
    "# Loading in networks from atlas file\n",
    "# Here we use first network in atlas, network_0 for example\n",
    "network = grn.GRN()\n",
    "network.load_dict(\n",
    "    json.decode(path = 'full_atlas.js')['network_0']\n",
    ")"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Then, a plotter object can be initialized as:"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "plotter = ageas.Plot(network = network,)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "And graph of query network can be plotted and saved:"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "plotter.draw(\n",
    "    legend_pos = (1.05, 0.3), # adjust legend position\n",
    "    edge_width_scale = 1.0,   # adjust edge(GRP) width\n",
    ")\n",
    "plotter.save(path = 'test_plot.png', format = 'png')"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Now, we should be able to get a network graph looks like: ![this one](../images/test.png)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Also, we can visualize GRPs by correlation strength calculated with class 1 samples or class 2 samples."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "plotter = ageas.Plot(\n",
    "    network = network,\n",
    "    plot_class = 'class2',  # using mef data here. 'class1' will be ipsc data\n",
    "    hide_bridge = False,    # unhide GRPs linking regulatory sources \n",
    ")\n",
    "plotter.draw(\n",
    "    legend_pos = (1.15, 0.3), # adjust legend position\n",
    "    edge_width_scale = 1.0,   # adjust edge(GRP) width\n",
    ")\n",
    "plotter.save(path = 'test_plot.png', format = 'png')"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Output should looks like: ![next one](../images/class2_plot.png)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Surely, we can only visualize GRPs related to one specific gene of interest in order to avoid over-crowded graph."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "plotter = ageas.Plot(\n",
    "    network = network,\n",
    "    hide_bridge = False,    # unhide GRPs linking regulatory sources\n",
    "    root_gene = 'Klf4',     # Klf4 is our gene of interest\n",
    "    depth = 1,              # and we are visualizing all GRPs directly linked with Klf4\n",
    ")\n",
    "plotter.draw(\n",
    "    figure_size = 10,           # less GRPs here, smaller size would be better\n",
    "    legend_pos = (1.3, 0.3),   # adjust legend position\n",
    "    edge_width_scale = 2.0,     # adjust edge(GRP) width\n",
    ")\n",
    "plotter.save(path = 'test_plot_1.png', format = 'png')"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "And we shoud get something looks like: ![klf4 one](../images/klf4.png)"
   ]
  }
 ],
 "metadata": {
  "kernelspec": {
   "display_name": "Python 3.9.13 64-bit (windows store)",
   "language": "python",
   "name": "python3"
  },
  "language_info": {
   "codemirror_mode": {
    "name": "ipython",
    "version": 3
   },
   "file_extension": ".py",
   "mimetype": "text/x-python",
   "name": "python",
   "nbconvert_exporter": "python",
   "pygments_lexer": "ipython3",
   "version": "3.9.13"
  },
  "orig_nbformat": 4,
  "vscode": {
   "interpreter": {
    "hash": "2a9e64e061ab733f2e33056f37bf3f62a8dd02da99810729dff6b17cfb3a5e9f"
   }
  }
 },
 "nbformat": 4,
 "nbformat_minor": 2
}