{
 "cells": [
  {
   "cell_type": "markdown",
   "id": "e4d3a1ad-1f33-49b4-b0df-917f13f1f820",
   "metadata": {},
   "source": [
    "# Genome Classifier "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 1,
   "id": "472aa6e1-c9aa-4d43-914e-4f71d3e6161c",
   "metadata": {},
   "outputs": [],
   "source": [
    "import modelseedpy"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "11c16251-0cdc-4cd8-a6be-1be80646677d",
   "metadata": {},
   "source": [
    "#### Pull the genome classifier model"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "id": "32587da5-317c-4a75-b3b1-c2467086a685",
   "metadata": {},
   "outputs": [],
   "source": [
    "from modelseedpy.helpers import get_classifier"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "id": "00c68e84-def5-446b-8bfa-311909287eab",
   "metadata": {},
   "outputs": [],
   "source": [
    "classifier = get_classifier('knn_filter')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "id": "eaef7784-140a-4e11-970a-b29cf023f5c9",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "modelseedpy.core.msgenomeclassifier.MSGenomeClassifier"
      ]
     },
     "execution_count": 4,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "type(classifier)"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "7e282aa9-f1b2-4a37-bf1e-25e29ef97480",
   "metadata": {},
   "source": [
    "#### Get a Genome and Annotate with RAST\n",
    "RAST annotation is essential since the classifier was trained with RAST annotated functions"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "id": "f4127633-a524-446b-9e8f-983ccf367035",
   "metadata": {},
   "outputs": [],
   "source": [
    "# Load e. coli genome\n",
    "genome = modelseedpy.MSGenome.from_fasta('GCF_000005845.2_ASM584v2_protein.faa', split=' ')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 6,
   "id": "73e8708a-ff54-4a6b-b777-adccfecc0192",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "[{'id': '23AFF380-F4F9-11EB-BBBA-BBE5BBF382BD',\n",
       "  'parameters': ['-a',\n",
       "   '-g',\n",
       "   200,\n",
       "   '-m',\n",
       "   5,\n",
       "   '-d',\n",
       "   '/opt/patric-common/data/kmer_metadata_v2',\n",
       "   '-u',\n",
       "   'http://pear.mcs.anl.gov:6100/query'],\n",
       "  'hostname': 'pear',\n",
       "  'tool_name': 'kmer_search',\n",
       "  'execution_time': 1628063644.76991},\n",
       " {'execution_time': 1628063644.90382,\n",
       "  'tool_name': 'KmerAnnotationByFigfam',\n",
       "  'hostname': 'pear',\n",
       "  'id': '23C46324-F4F9-11EB-BBBA-BBE5BBF382BD',\n",
       "  'parameters': ['annotate_hypothetical_only=1',\n",
       "   'dataset_name=Release70',\n",
       "   'kmer_size=8']},\n",
       " {'parameters': [],\n",
       "  'id': '23F64B78-F4F9-11EB-908D-F73BBDF382BD',\n",
       "  'tool_name': 'annotate_proteins_similarity',\n",
       "  'hostname': 'pear',\n",
       "  'execute_time': 1628063645.23091}]"
      ]
     },
     "execution_count": 6,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "modelseedpy.RastClient().annotate_genome(genome)"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "6aa0fe52-eb01-4bee-9fd1-c5125c1e5f5e",
   "metadata": {},
   "source": [
    "#### Run classifier\n",
    "- A: Archaea\n",
    "- C: Cyanobacteria\n",
    "- N: Gram Negative\n",
    "- P: Gram Positive"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "ab9ae991-c4d7-494d-8cd5-f6923c534c8a",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "'N'"
      ]
     },
     "execution_count": 7,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "classifier.classify(genome)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "ba6b6bf2-936b-4e84-a90f-117d2c294942",
   "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": "3.9.12"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 5
}