Example(gff).ipynb

{
 "cells": [
  {
   "cell_type": "markdown",
   "source": [
    "# **Welcome to Jupyter!**\r\n",
    "## To load the pandas package and start working with it, import the package. The community agreed alias for pandas is pd, so loading pandas as pd is assumed standard practice for all of the pandas documentation."
   ],
   "metadata": {}
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "source": [
    "import pandas as pd\r\n",
    "import matplotlib.pyplot as plt"
   ],
   "outputs": [],
   "metadata": {}
  },
  {
   "cell_type": "markdown",
   "source": [
    "## I imported matplotlib.pyplot as plt for another reason that I will explain later"
   ],
   "metadata": {}
  },
  {
   "cell_type": "markdown",
   "source": [
    "## pandas supports the integration with many file formats or data sources out of the box (csv, excel, sql, json, parquet,…).\r\n",
    " ## Importing data from each of these data sources is provided by function with the prefix *read_* Similarly, the *to_* methods are used to store data. \r\n",
    "\r\n",
    "## We will practice by loading a GFF file. We provide the argument comment=\"#\" below so it will skip the first few lines of the file that are not data but just comments, and we will enter in the column names by hand because it is easier than trying to parse them from the file in this case."
   ],
   "metadata": {}
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "source": [
    "# read csv function with arguments to parse the file\r\n",
    "gff = pd.read_csv(\r\n",
    "    \"./NZ_GG770409.gff\", \r\n",
    "    sep=\"\\t\", \r\n",
    "    header=None,\r\n",
    "    comment=\"#\", \r\n",
    "    names=(\"seqname\", \"source\", \"feature\", \"start\",\r\n",
    "           \"end\", \"score\", \"strand\", \"frame\", \"attribute\"),\r\n",
    ")"
   ],
   "outputs": [],
   "metadata": {}
  },
  {
   "cell_type": "markdown",
   "source": [
    "## Make sure to always have a check on the data after reading it.\r\n",
    "## When displaying a DataFrame, the first 5 rows will be shown by default:\r\n"
   ],
   "metadata": {}
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "source": [
    "gff.head(8)"
   ],
   "outputs": [
    {
     "output_type": "execute_result",
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>seqname</th>\n",
       "      <th>source</th>\n",
       "      <th>feature</th>\n",
       "      <th>start</th>\n",
       "      <th>end</th>\n",
       "      <th>score</th>\n",
       "      <th>strand</th>\n",
       "      <th>frame</th>\n",
       "      <th>attribute</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>NZ_GG770409.1</td>\n",
       "      <td>RefSeq</td>\n",
       "      <td>region</td>\n",
       "      <td>1</td>\n",
       "      <td>6115515</td>\n",
       "      <td>.</td>\n",
       "      <td>+</td>\n",
       "      <td>.</td>\n",
       "      <td>ID=id0;Dbxref=Greengenes:253191,HMP:0004,taxon...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>NZ_GG770409.1</td>\n",
       "      <td>RefSeq</td>\n",
       "      <td>gene</td>\n",
       "      <td>1</td>\n",
       "      <td>288</td>\n",
       "      <td>.</td>\n",
       "      <td>-</td>\n",
       "      <td>.</td>\n",
       "      <td>ID=gene0;Name=HMPREF0004_0003;gbkey=Gene;locus...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>NZ_GG770409.1</td>\n",
       "      <td>RefSeq</td>\n",
       "      <td>CDS</td>\n",
       "      <td>1</td>\n",
       "      <td>288</td>\n",
       "      <td>.</td>\n",
       "      <td>-</td>\n",
       "      <td>0</td>\n",
       "      <td>ID=cds0;Name=ZP_06684427.1;Parent=gene0;Dbxref...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>NZ_GG770409.1</td>\n",
       "      <td>RefSeq</td>\n",
       "      <td>gene</td>\n",
       "      <td>316</td>\n",
       "      <td>1767</td>\n",
       "      <td>.</td>\n",
       "      <td>-</td>\n",
       "      <td>.</td>\n",
       "      <td>ID=gene1;Name=HMPREF0004_0004;gbkey=Gene;locus...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>NZ_GG770409.1</td>\n",
       "      <td>RefSeq</td>\n",
       "      <td>CDS</td>\n",
       "      <td>316</td>\n",
       "      <td>1767</td>\n",
       "      <td>.</td>\n",
       "      <td>-</td>\n",
       "      <td>0</td>\n",
       "      <td>ID=cds1;Name=ZP_06684428.1;Parent=gene1;Note=C...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>5</th>\n",
       "      <td>NZ_GG770409.1</td>\n",
       "      <td>RefSeq</td>\n",
       "      <td>gene</td>\n",
       "      <td>1885</td>\n",
       "      <td>4315</td>\n",
       "      <td>.</td>\n",
       "      <td>-</td>\n",
       "      <td>.</td>\n",
       "      <td>ID=gene2;Name=HMPREF0004_0005;gbkey=Gene;locus...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>6</th>\n",
       "      <td>NZ_GG770409.1</td>\n",
       "      <td>RefSeq</td>\n",
       "      <td>CDS</td>\n",
       "      <td>1885</td>\n",
       "      <td>4315</td>\n",
       "      <td>.</td>\n",
       "      <td>-</td>\n",
       "      <td>0</td>\n",
       "      <td>ID=cds2;Name=ZP_06684429.1;Parent=gene2;Note=C...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>7</th>\n",
       "      <td>NZ_GG770409.1</td>\n",
       "      <td>RefSeq</td>\n",
       "      <td>gene</td>\n",
       "      <td>4408</td>\n",
       "      <td>5016</td>\n",
       "      <td>.</td>\n",
       "      <td>-</td>\n",
       "      <td>.</td>\n",
       "      <td>ID=gene3;Name=HMPREF0004_0006;gbkey=Gene;locus...</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "         seqname  source feature  start      end score strand frame  \\\n",
       "0  NZ_GG770409.1  RefSeq  region      1  6115515     .      +     .   \n",
       "1  NZ_GG770409.1  RefSeq    gene      1      288     .      -     .   \n",
       "2  NZ_GG770409.1  RefSeq     CDS      1      288     .      -     0   \n",
       "3  NZ_GG770409.1  RefSeq    gene    316     1767     .      -     .   \n",
       "4  NZ_GG770409.1  RefSeq     CDS    316     1767     .      -     0   \n",
       "5  NZ_GG770409.1  RefSeq    gene   1885     4315     .      -     .   \n",
       "6  NZ_GG770409.1  RefSeq     CDS   1885     4315     .      -     0   \n",
       "7  NZ_GG770409.1  RefSeq    gene   4408     5016     .      -     .   \n",
       "\n",
       "                                           attribute  \n",
       "0  ID=id0;Dbxref=Greengenes:253191,HMP:0004,taxon...  \n",
       "1  ID=gene0;Name=HMPREF0004_0003;gbkey=Gene;locus...  \n",
       "2  ID=cds0;Name=ZP_06684427.1;Parent=gene0;Dbxref...  \n",
       "3  ID=gene1;Name=HMPREF0004_0004;gbkey=Gene;locus...  \n",
       "4  ID=cds1;Name=ZP_06684428.1;Parent=gene1;Note=C...  \n",
       "5  ID=gene2;Name=HMPREF0004_0005;gbkey=Gene;locus...  \n",
       "6  ID=cds2;Name=ZP_06684429.1;Parent=gene2;Note=C...  \n",
       "7  ID=gene3;Name=HMPREF0004_0006;gbkey=Gene;locus...  "
      ]
     },
     "metadata": {},
     "execution_count": 4
    }
   ],
   "metadata": {}
  },
  {
   "cell_type": "markdown",
   "source": [
    "## Use the head() method with the required number of rows (in this case 8) as an argument to see the first N rows of a DataFrame (the default is 5).\r\n",
    "## If you want to see the last N rows, use the tail() method.\r\n",
    "## To see the last two rows, what should we type in the next cell?"
   ],
   "metadata": {}
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "source": [],
   "outputs": [],
   "metadata": {}
  },
  {
   "cell_type": "markdown",
   "source": [
    "## You can check how pandas interpreted each of the column data types  by dtypes attribute."
   ],
   "metadata": {}
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "source": [
    "gff.dtypes"
   ],
   "outputs": [
    {
     "output_type": "execute_result",
     "data": {
      "text/plain": [
       "seqname      object\n",
       "source       object\n",
       "feature      object\n",
       "start         int64\n",
       "end           int64\n",
       "score        object\n",
       "strand       object\n",
       "frame        object\n",
       "attribute    object\n",
       "dtype: object"
      ]
     },
     "metadata": {},
     "execution_count": 5
    }
   ],
   "metadata": {}
  },
  {
   "cell_type": "markdown",
   "source": [
    "## If you are Interested in a technical summary of the DataFrame, use info()."
   ],
   "metadata": {}
  },
  {
   "cell_type": "code",
   "execution_count": 6,
   "source": [
    "gff.info()"
   ],
   "outputs": [
    {
     "output_type": "stream",
     "name": "stdout",
     "text": [
      "<class 'pandas.core.frame.DataFrame'>\n",
      "RangeIndex: 11591 entries, 0 to 11590\n",
      "Data columns (total 9 columns):\n",
      " #   Column     Non-Null Count  Dtype \n",
      "---  ------     --------------  ----- \n",
      " 0   seqname    11591 non-null  object\n",
      " 1   source     11591 non-null  object\n",
      " 2   feature    11591 non-null  object\n",
      " 3   start      11591 non-null  int64 \n",
      " 4   end        11591 non-null  int64 \n",
      " 5   score      11591 non-null  object\n",
      " 6   strand     11591 non-null  object\n",
      " 7   frame      11591 non-null  object\n",
      " 8   attribute  11591 non-null  object\n",
      "dtypes: int64(2), object(7)\n",
      "memory usage: 815.1+ KB\n"
     ]
    }
   ],
   "metadata": {}
  },
  {
   "cell_type": "markdown",
   "source": [
    "\r\n",
    "## Some of the columns may pique your interest at times. You can select a (some) specific column (s); for example, if you want to see the 'feature' column, you can do so as follows:"
   ],
   "metadata": {}
  },
  {
   "cell_type": "code",
   "execution_count": 7,
   "source": [
    "features = gff['feature']\r\n",
    "features.head()"
   ],
   "outputs": [
    {
     "output_type": "execute_result",
     "data": {
      "text/plain": [
       "0    region\n",
       "1      gene\n",
       "2       CDS\n",
       "3      gene\n",
       "4       CDS\n",
       "Name: feature, dtype: object"
      ]
     },
     "metadata": {},
     "execution_count": 7
    }
   ],
   "metadata": {}
  },
  {
   "cell_type": "markdown",
   "source": [
    "##  A convenient way to filter or subselect part of a data set is to use a mask. You can create a boolean array to act as a mask for an entire column or row by asking a question with a True or False answer and broadcasting it across the row or column. For example, below we ask whether the feature is a CDS, which returns a True or False for every element in the 'feature' column."
   ],
   "metadata": {}
  },
  {
   "cell_type": "code",
   "execution_count": 8,
   "source": [
    "# create a boolean mask of the feature column asking 'is it a CDS?'\r\n",
    "mask = gff.feature == \"CDS\"\r\n",
    "\r\n",
    "# show the first 10 values\r\n",
    "mask.head(10)"
   ],
   "outputs": [
    {
     "output_type": "execute_result",
     "data": {
      "text/plain": [
       "0    False\n",
       "1    False\n",
       "2     True\n",
       "3    False\n",
       "4     True\n",
       "5    False\n",
       "6     True\n",
       "7    False\n",
       "8     True\n",
       "9    False\n",
       "Name: feature, dtype: bool"
      ]
     },
     "metadata": {},
     "execution_count": 8
    }
   ],
   "metadata": {}
  },
  {
   "cell_type": "markdown",
   "source": [
    "## We can now apply the mask to the DataFrame as an index, or with.loc or.iloc, and it will select all rows with True, but not those with False. The index (row name) on the right is kept across the entire dataframe, so if you select only CDS, you'll get element 2, 4, 6, 8, and so on."
   ],
   "metadata": {}
  },
  {
   "cell_type": "code",
   "execution_count": 9,
   "source": [
    "# use this mask to select only rows that are CDS\r\n",
    "CDSs = gff[mask]\r\n",
    "\r\n",
    "# return the first 10 value\r\n",
    "CDSs.head()"
   ],
   "outputs": [
    {
     "output_type": "execute_result",
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>seqname</th>\n",
       "      <th>source</th>\n",
       "      <th>feature</th>\n",
       "      <th>start</th>\n",
       "      <th>end</th>\n",
       "      <th>score</th>\n",
       "      <th>strand</th>\n",
       "      <th>frame</th>\n",
       "      <th>attribute</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>NZ_GG770409.1</td>\n",
       "      <td>RefSeq</td>\n",
       "      <td>CDS</td>\n",
       "      <td>1</td>\n",
       "      <td>288</td>\n",
       "      <td>.</td>\n",
       "      <td>-</td>\n",
       "      <td>0</td>\n",
       "      <td>ID=cds0;Name=ZP_06684427.1;Parent=gene0;Dbxref...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>NZ_GG770409.1</td>\n",
       "      <td>RefSeq</td>\n",
       "      <td>CDS</td>\n",
       "      <td>316</td>\n",
       "      <td>1767</td>\n",
       "      <td>.</td>\n",
       "      <td>-</td>\n",
       "      <td>0</td>\n",
       "      <td>ID=cds1;Name=ZP_06684428.1;Parent=gene1;Note=C...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>6</th>\n",
       "      <td>NZ_GG770409.1</td>\n",
       "      <td>RefSeq</td>\n",
       "      <td>CDS</td>\n",
       "      <td>1885</td>\n",
       "      <td>4315</td>\n",
       "      <td>.</td>\n",
       "      <td>-</td>\n",
       "      <td>0</td>\n",
       "      <td>ID=cds2;Name=ZP_06684429.1;Parent=gene2;Note=C...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>8</th>\n",
       "      <td>NZ_GG770409.1</td>\n",
       "      <td>RefSeq</td>\n",
       "      <td>CDS</td>\n",
       "      <td>4408</td>\n",
       "      <td>5016</td>\n",
       "      <td>.</td>\n",
       "      <td>-</td>\n",
       "      <td>0</td>\n",
       "      <td>ID=cds3;Name=ZP_06684430.1;Parent=gene3;Note=C...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>10</th>\n",
       "      <td>NZ_GG770409.1</td>\n",
       "      <td>RefSeq</td>\n",
       "      <td>CDS</td>\n",
       "      <td>5013</td>\n",
       "      <td>5294</td>\n",
       "      <td>.</td>\n",
       "      <td>-</td>\n",
       "      <td>0</td>\n",
       "      <td>ID=cds4;Name=ZP_06684431.1;Parent=gene4;Note=I...</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "          seqname  source feature  start   end score strand frame  \\\n",
       "2   NZ_GG770409.1  RefSeq     CDS      1   288     .      -     0   \n",
       "4   NZ_GG770409.1  RefSeq     CDS    316  1767     .      -     0   \n",
       "6   NZ_GG770409.1  RefSeq     CDS   1885  4315     .      -     0   \n",
       "8   NZ_GG770409.1  RefSeq     CDS   4408  5016     .      -     0   \n",
       "10  NZ_GG770409.1  RefSeq     CDS   5013  5294     .      -     0   \n",
       "\n",
       "                                            attribute  \n",
       "2   ID=cds0;Name=ZP_06684427.1;Parent=gene0;Dbxref...  \n",
       "4   ID=cds1;Name=ZP_06684428.1;Parent=gene1;Note=C...  \n",
       "6   ID=cds2;Name=ZP_06684429.1;Parent=gene2;Note=C...  \n",
       "8   ID=cds3;Name=ZP_06684430.1;Parent=gene3;Note=C...  \n",
       "10  ID=cds4;Name=ZP_06684431.1;Parent=gene4;Note=I...  "
      ]
     },
     "metadata": {},
     "execution_count": 9
    }
   ],
   "metadata": {}
  },
  {
   "cell_type": "code",
   "execution_count": 11,
   "source": [
    " # or, we could have written the above code more concisely as a single line:\r\n",
    "gff.loc[gff.feature == \"CDS\"].head()"
   ],
   "outputs": [
    {
     "output_type": "execute_result",
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>seqname</th>\n",
       "      <th>source</th>\n",
       "      <th>feature</th>\n",
       "      <th>start</th>\n",
       "      <th>end</th>\n",
       "      <th>score</th>\n",
       "      <th>strand</th>\n",
       "      <th>frame</th>\n",
       "      <th>attribute</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>NZ_GG770409.1</td>\n",
       "      <td>RefSeq</td>\n",
       "      <td>CDS</td>\n",
       "      <td>1</td>\n",
       "      <td>288</td>\n",
       "      <td>.</td>\n",
       "      <td>-</td>\n",
       "      <td>0</td>\n",
       "      <td>ID=cds0;Name=ZP_06684427.1;Parent=gene0;Dbxref...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>NZ_GG770409.1</td>\n",
       "      <td>RefSeq</td>\n",
       "      <td>CDS</td>\n",
       "      <td>316</td>\n",
       "      <td>1767</td>\n",
       "      <td>.</td>\n",
       "      <td>-</td>\n",
       "      <td>0</td>\n",
       "      <td>ID=cds1;Name=ZP_06684428.1;Parent=gene1;Note=C...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>6</th>\n",
       "      <td>NZ_GG770409.1</td>\n",
       "      <td>RefSeq</td>\n",
       "      <td>CDS</td>\n",
       "      <td>1885</td>\n",
       "      <td>4315</td>\n",
       "      <td>.</td>\n",
       "      <td>-</td>\n",
       "      <td>0</td>\n",
       "      <td>ID=cds2;Name=ZP_06684429.1;Parent=gene2;Note=C...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>8</th>\n",
       "      <td>NZ_GG770409.1</td>\n",
       "      <td>RefSeq</td>\n",
       "      <td>CDS</td>\n",
       "      <td>4408</td>\n",
       "      <td>5016</td>\n",
       "      <td>.</td>\n",
       "      <td>-</td>\n",
       "      <td>0</td>\n",
       "      <td>ID=cds3;Name=ZP_06684430.1;Parent=gene3;Note=C...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>10</th>\n",
       "      <td>NZ_GG770409.1</td>\n",
       "      <td>RefSeq</td>\n",
       "      <td>CDS</td>\n",
       "      <td>5013</td>\n",
       "      <td>5294</td>\n",
       "      <td>.</td>\n",
       "      <td>-</td>\n",
       "      <td>0</td>\n",
       "      <td>ID=cds4;Name=ZP_06684431.1;Parent=gene4;Note=I...</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "          seqname  source feature  start   end score strand frame  \\\n",
       "2   NZ_GG770409.1  RefSeq     CDS      1   288     .      -     0   \n",
       "4   NZ_GG770409.1  RefSeq     CDS    316  1767     .      -     0   \n",
       "6   NZ_GG770409.1  RefSeq     CDS   1885  4315     .      -     0   \n",
       "8   NZ_GG770409.1  RefSeq     CDS   4408  5016     .      -     0   \n",
       "10  NZ_GG770409.1  RefSeq     CDS   5013  5294     .      -     0   \n",
       "\n",
       "                                            attribute  \n",
       "2   ID=cds0;Name=ZP_06684427.1;Parent=gene0;Dbxref...  \n",
       "4   ID=cds1;Name=ZP_06684428.1;Parent=gene1;Note=C...  \n",
       "6   ID=cds2;Name=ZP_06684429.1;Parent=gene2;Note=C...  \n",
       "8   ID=cds3;Name=ZP_06684430.1;Parent=gene3;Note=C...  \n",
       "10  ID=cds4;Name=ZP_06684431.1;Parent=gene4;Note=I...  "
      ]
     },
     "metadata": {},
     "execution_count": 11
    }
   ],
   "metadata": {}
  },
  {
   "cell_type": "markdown",
   "source": [
    "## In the code cells below we are just returning the filtered DataDrame rather than saving it to a variable so you can easily see the result of the operation. As you can see it is easy to explore the effect of our operations by filtering in various ways and calling the .head() function to show a small part of the result."
   ],
   "metadata": {}
  },
  {
   "cell_type": "code",
   "execution_count": 12,
   "source": [
    "gff[gff.feature == \"CDS\"].head()"
   ],
   "outputs": [
    {
     "output_type": "execute_result",
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>seqname</th>\n",
       "      <th>source</th>\n",
       "      <th>feature</th>\n",
       "      <th>start</th>\n",
       "      <th>end</th>\n",
       "      <th>score</th>\n",
       "      <th>strand</th>\n",
       "      <th>frame</th>\n",
       "      <th>attribute</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>NZ_GG770409.1</td>\n",
       "      <td>RefSeq</td>\n",
       "      <td>CDS</td>\n",
       "      <td>1</td>\n",
       "      <td>288</td>\n",
       "      <td>.</td>\n",
       "      <td>-</td>\n",
       "      <td>0</td>\n",
       "      <td>ID=cds0;Name=ZP_06684427.1;Parent=gene0;Dbxref...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>NZ_GG770409.1</td>\n",
       "      <td>RefSeq</td>\n",
       "      <td>CDS</td>\n",
       "      <td>316</td>\n",
       "      <td>1767</td>\n",
       "      <td>.</td>\n",
       "      <td>-</td>\n",
       "      <td>0</td>\n",
       "      <td>ID=cds1;Name=ZP_06684428.1;Parent=gene1;Note=C...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>6</th>\n",
       "      <td>NZ_GG770409.1</td>\n",
       "      <td>RefSeq</td>\n",
       "      <td>CDS</td>\n",
       "      <td>1885</td>\n",
       "      <td>4315</td>\n",
       "      <td>.</td>\n",
       "      <td>-</td>\n",
       "      <td>0</td>\n",
       "      <td>ID=cds2;Name=ZP_06684429.1;Parent=gene2;Note=C...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>8</th>\n",
       "      <td>NZ_GG770409.1</td>\n",
       "      <td>RefSeq</td>\n",
       "      <td>CDS</td>\n",
       "      <td>4408</td>\n",
       "      <td>5016</td>\n",
       "      <td>.</td>\n",
       "      <td>-</td>\n",
       "      <td>0</td>\n",
       "      <td>ID=cds3;Name=ZP_06684430.1;Parent=gene3;Note=C...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>10</th>\n",
       "      <td>NZ_GG770409.1</td>\n",
       "      <td>RefSeq</td>\n",
       "      <td>CDS</td>\n",
       "      <td>5013</td>\n",
       "      <td>5294</td>\n",
       "      <td>.</td>\n",
       "      <td>-</td>\n",
       "      <td>0</td>\n",
       "      <td>ID=cds4;Name=ZP_06684431.1;Parent=gene4;Note=I...</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "          seqname  source feature  start   end score strand frame  \\\n",
       "2   NZ_GG770409.1  RefSeq     CDS      1   288     .      -     0   \n",
       "4   NZ_GG770409.1  RefSeq     CDS    316  1767     .      -     0   \n",
       "6   NZ_GG770409.1  RefSeq     CDS   1885  4315     .      -     0   \n",
       "8   NZ_GG770409.1  RefSeq     CDS   4408  5016     .      -     0   \n",
       "10  NZ_GG770409.1  RefSeq     CDS   5013  5294     .      -     0   \n",
       "\n",
       "                                            attribute  \n",
       "2   ID=cds0;Name=ZP_06684427.1;Parent=gene0;Dbxref...  \n",
       "4   ID=cds1;Name=ZP_06684428.1;Parent=gene1;Note=C...  \n",
       "6   ID=cds2;Name=ZP_06684429.1;Parent=gene2;Note=C...  \n",
       "8   ID=cds3;Name=ZP_06684430.1;Parent=gene3;Note=C...  \n",
       "10  ID=cds4;Name=ZP_06684431.1;Parent=gene4;Note=I...  "
      ]
     },
     "metadata": {},
     "execution_count": 12
    }
   ],
   "metadata": {}
  },
  {
   "cell_type": "code",
   "execution_count": 13,
   "source": [
    "# select attributes of those rows\r\n",
    "gff[gff.feature == \"CDS\"][\"attribute\"].head(5)"
   ],
   "outputs": [
    {
     "output_type": "execute_result",
     "data": {
      "text/plain": [
       "2     ID=cds0;Name=ZP_06684427.1;Parent=gene0;Dbxref...\n",
       "4     ID=cds1;Name=ZP_06684428.1;Parent=gene1;Note=C...\n",
       "6     ID=cds2;Name=ZP_06684429.1;Parent=gene2;Note=C...\n",
       "8     ID=cds3;Name=ZP_06684430.1;Parent=gene3;Note=C...\n",
       "10    ID=cds4;Name=ZP_06684431.1;Parent=gene4;Note=I...\n",
       "Name: attribute, dtype: object"
      ]
     },
     "metadata": {},
     "execution_count": 13
    }
   ],
   "metadata": {}
  },
  {
   "cell_type": "markdown",
   "source": [
    " ##  you can count the number of genes in the DataFrame in the code cell below."
   ],
   "metadata": {}
  },
  {
   "cell_type": "code",
   "execution_count": 15,
   "source": [
    "gff.loc[gff.feature == \"gene\"].shape[0]"
   ],
   "outputs": [
    {
     "output_type": "execute_result",
     "data": {
      "text/plain": [
       "5761"
      ]
     },
     "metadata": {},
     "execution_count": 15
    }
   ],
   "metadata": {}
  },
  {
   "cell_type": "markdown",
   "source": [
    "## We can also use the iloc[] attribute to specify the position of the passed integers. Because iloc is integer position-based, you must specify rows and columns by their integer position values (0-based integer position).\r\n",
    "\r\n",
    "\r\n",
    "## loc is label-based, which means that you have to specify rows and columns based on their row and column labels."
   ],
   "metadata": {}
  },
  {
   "cell_type": "code",
   "execution_count": 23,
   "source": [
    "#gff.iloc[rowindex, column index]\r\n",
    "gff.iloc[2:5, 3:7]"
   ],
   "outputs": [
    {
     "output_type": "execute_result",
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>start</th>\n",
       "      <th>end</th>\n",
       "      <th>score</th>\n",
       "      <th>strand</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>1</td>\n",
       "      <td>288</td>\n",
       "      <td>.</td>\n",
       "      <td>-</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>316</td>\n",
       "      <td>1767</td>\n",
       "      <td>.</td>\n",
       "      <td>-</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>316</td>\n",
       "      <td>1767</td>\n",
       "      <td>.</td>\n",
       "      <td>-</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "   start   end score strand\n",
       "2      1   288     .      -\n",
       "3    316  1767     .      -\n",
       "4    316  1767     .      -"
      ]
     },
     "metadata": {},
     "execution_count": 23
    }
   ],
   "metadata": {}
  },
  {
   "cell_type": "markdown",
   "source": [
    "## What should I write if I want rows 2 to 6 and all the columns?"
   ],
   "metadata": {}
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "source": [],
   "outputs": [],
   "metadata": {}
  },
  {
   "cell_type": "code",
   "execution_count": 16,
   "source": [
    "# 3 is the label of row 4\r\n",
    "gff.loc[3]"
   ],
   "outputs": [
    {
     "output_type": "execute_result",
     "data": {
      "text/plain": [
       "seqname                                          NZ_GG770409.1\n",
       "source                                                  RefSeq\n",
       "feature                                                   gene\n",
       "start                                                      316\n",
       "end                                                       1767\n",
       "score                                                        .\n",
       "strand                                                       -\n",
       "frame                                                        .\n",
       "attribute    ID=gene1;Name=HMPREF0004_0004;gbkey=Gene;locus...\n",
       "Name: 3, dtype: object"
      ]
     },
     "metadata": {},
     "execution_count": 16
    }
   ],
   "metadata": {}
  },
  {
   "cell_type": "markdown",
   "source": [
    "## A DataFrame can be filtered to show only specific rows. For example, if you're looking for samples with a starting position greater than 4408:\r\n"
   ],
   "metadata": {}
  },
  {
   "cell_type": "code",
   "execution_count": 17,
   "source": [
    "above_4408 = gff[gff['start']> 4408]\r\n",
    "above_4408.head(3)"
   ],
   "outputs": [
    {
     "output_type": "execute_result",
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>seqname</th>\n",
       "      <th>source</th>\n",
       "      <th>feature</th>\n",
       "      <th>start</th>\n",
       "      <th>end</th>\n",
       "      <th>score</th>\n",
       "      <th>strand</th>\n",
       "      <th>frame</th>\n",
       "      <th>attribute</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>9</th>\n",
       "      <td>NZ_GG770409.1</td>\n",
       "      <td>RefSeq</td>\n",
       "      <td>gene</td>\n",
       "      <td>5013</td>\n",
       "      <td>5294</td>\n",
       "      <td>.</td>\n",
       "      <td>-</td>\n",
       "      <td>.</td>\n",
       "      <td>ID=gene4;Name=HMPREF0004_0007;gbkey=Gene;locus...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>10</th>\n",
       "      <td>NZ_GG770409.1</td>\n",
       "      <td>RefSeq</td>\n",
       "      <td>CDS</td>\n",
       "      <td>5013</td>\n",
       "      <td>5294</td>\n",
       "      <td>.</td>\n",
       "      <td>-</td>\n",
       "      <td>0</td>\n",
       "      <td>ID=cds4;Name=ZP_06684431.1;Parent=gene4;Note=I...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>11</th>\n",
       "      <td>NZ_GG770409.1</td>\n",
       "      <td>RefSeq</td>\n",
       "      <td>gene</td>\n",
       "      <td>5569</td>\n",
       "      <td>5838</td>\n",
       "      <td>.</td>\n",
       "      <td>+</td>\n",
       "      <td>.</td>\n",
       "      <td>ID=gene5;Name=coxA;gbkey=Gene;gene=coxA;locus_...</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "          seqname  source feature  start   end score strand frame  \\\n",
       "9   NZ_GG770409.1  RefSeq    gene   5013  5294     .      -     .   \n",
       "10  NZ_GG770409.1  RefSeq     CDS   5013  5294     .      -     0   \n",
       "11  NZ_GG770409.1  RefSeq    gene   5569  5838     .      +     .   \n",
       "\n",
       "                                            attribute  \n",
       "9   ID=gene4;Name=HMPREF0004_0007;gbkey=Gene;locus...  \n",
       "10  ID=cds4;Name=ZP_06684431.1;Parent=gene4;Note=I...  \n",
       "11  ID=gene5;Name=coxA;gbkey=Gene;gene=coxA;locus_...  "
      ]
     },
     "metadata": {},
     "execution_count": 17
    }
   ],
   "metadata": {}
  },
  {
   "cell_type": "markdown",
   "source": [
    "## Pandas plots your data using Matplotlib's power. You can select the appropriate plot type (scatter, bar, boxplot, etc.) based on your data.\r\n",
    "## I've brought two different plot types, but there are many more. If you want to see them, go to this [website](https://pandas.pydata.org/pandas-docs/stable/user_guide/visualization.html)"
   ],
   "metadata": {}
  },
  {
   "cell_type": "code",
   "execution_count": 18,
   "source": [
    "gff['start'].plot(kind='hist', bins = 50, color=\"b\")"
   ],
   "outputs": [
    {
     "output_type": "execute_result",
     "data": {
      "text/plain": [
       "<AxesSubplot:ylabel='Frequency'>"
      ]
     },
     "metadata": {},
     "execution_count": 18
    },
    {
     "output_type": "display_data",
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAYUAAAEFCAYAAAAMk/uQAAAAOXRFWHRTb2Z0d2FyZQBNYXRwbG90bGliIHZlcnNpb24zLjMuNCwgaHR0cHM6Ly9tYXRwbG90bGliLm9yZy8QVMy6AAAACXBIWXMAAAsTAAALEwEAmpwYAAASc0lEQVR4nO3da6xlZX3H8e+PS1FRo2YGOgJ20IwXMAp4pBda66Veq6JtsEOsIYaKTdFotK1ATLUvSHzhrU3VOgoVFcRRRGljVaDUS1OFM4jCMFAngnIcyoy3ANaA4L8v9prldjgzs8+ZWXvtffb3k5zstZ69197/lXP5nfWsZz0rVYUkSQAH9F2AJGlyGAqSpJahIElqGQqSpJahIElqHdR3Afti1apVtXbt2r7LkKSpsmnTph9W1erFnpvqUFi7di3z8/N9lyFJUyXJ93b3nN1HkqSWoSBJahkKkqSWoSBJahkKkqSWoSBJahkKkqSWoSBJahkKkqTWVF/RrMmTLN7uvZyk6eCRgiSpZShIklqGgiSpZShIklqGgiSpZShIklqGgiSpZShIklpevCaNmRf4aZJ5pCBJahkKkqSW3UdL4GH/bJuW7/+01KnJ5JGCJKllKEiSWoaCJKnVWSgkOSrJVUm2JNmc5A1N+9uT/CDJdc3Xi4a2OTvJ1iQ3J3l+V7VJ6kay+JemR5cnmu8D3lxV1yZ5GLApyeXNc++pqncOvzjJMcB64Fjg0cAVSR5fVfd3WKMkaUhnRwpVdXtVXdss3wVsAY7YwyYnAxdX1T1VdQuwFTixq/q0svgf6nTy+zZ5xnJOIcla4HjgG03T65J8O8n5SR7ZtB0B3Da02QKLhEiSM5LMJ5nfsWNHR/X6g6rZ4c+7hnUeCkkeClwCvLGq7gQ+ADwOOA64HXjXzpcusvkDRlZX1YaqmququdWrV3dTdE/85ZTUt05DIcnBDALhwqr6DEBV3VFV91fVL4EP8asuogXgqKHNjwS2dVlf1/wjL2nadDn6KMB5wJaqevdQ+5qhl70cuKFZvgxYn+SQJEcD64Cru6qvT4aFpEnV5eijk4BXAdcnua5pOwc4NclxDLqGbgVeC1BVm5NsBG5kMHLpTEceSdJ4dRYKVfU1Fj9P8Pk9bHMucG5XNXVlJfyX73w5AyvhezmLuv75naXfD69oliS1DAVJUsupsyVpmfbU3TitXUuGgiQ1PKdk95EkaYhHClNgqSMf9udIif31XrM0ekOaZoaCpJljN9HuGQpa0TxCkZbGUJBmxEr473gl7MOkMxQ0kfwPX/uDIbJ0Mx0K/sBI0q+b6VCQ9geParSSeJ2CJKllKEiSWnYfSdIEmJRuSENhhkziifVJrEmaJOMOC0NBU8UQmU5+36aHoSBJYzTpAemJZklSy1CQJLXsPtKyTPohsKTl8UhBktQyFCRJLbuPJE0cuyf7YyhoJvlHR12b1p8xu48kSS2PFKSOTOt/ippthoI04SZlojTNBruPJEmtzkIhyVFJrkqyJcnmJG9o2h+V5PIk32keHzm0zdlJtia5Ocnzu6pNmkTJ4l/SOHV5pHAf8OaqehLwO8CZSY4BzgKurKp1wJXNOs1z64FjgRcA709yYIf1SZJ20VkoVNXtVXVts3wXsAU4AjgZuKB52QXAy5rlk4GLq+qeqroF2Aqc2FV9kqQHGss5hSRrgeOBbwCHV9XtMAgO4LDmZUcAtw1tttC07fpeZySZTzK/Y8eOTuuWJpndTepC56GQ5KHAJcAbq+rOPb10kbYHjK+oqg1VNVdVc6tXr95fZUqS6DgUkhzMIBAurKrPNM13JFnTPL8G2N60LwBHDW1+JLCty/okSb+uy9FHAc4DtlTVu4eeugw4rVk+DfjcUPv6JIckORpYB1zdVX2SpAfq8uK1k4BXAdcnua5pOwd4B7AxyenA94FTAKpqc5KNwI0MRi6dWVX3d1ifJGkXnYVCVX2Nxc8TADxnN9ucC5zbVU0rjScVJe1vXtEsSWo595GkRe3pSNR5l1YujxQkSS1DQZLUsvtI0pI5yGHl8khBktQyFCRJLUNBktQyFCRJLUNBktQyFCRJLUNBktQaKRSSPLnrQiRJ/Rv1SOGfk1yd5K+SPKLLgiRJ/RkpFKrq94FXMrgz2nySi5I8t9PKJEljN/I5har6DvBW4C3AHwL/mOSmJH/SVXGSpPEa9ZzCU5K8B9gCPBt4SVU9qVl+T4f1SZLGaNQJ8f4J+BBwTlX9fGdjVW1L8tZOKpMkjd2oofAi4Oc775mc5ADgQVX1f1X1sc6qkySN1ajnFK4AHjy0/pCmTZK0gowaCg+qqrt3rjTLD+mmJElSX0YNhZ8lOWHnSpKnAT/fw+slSVNo1HMKbwQ+lWRbs74G+LNOKpIk9WakUKiqa5I8EXgCEOCmqvpFp5VJksZuKfdofjqwttnm+CRU1Uc7qUqS1IuRQiHJx4DHAdcB9zfNBRgKkrSCjHqkMAccU1XVZTGSpH6NOvroBuA3uyxEktS/UUNhFXBjki8muWzn1542SHJ+ku1Jbhhqe3uSHyS5rvl60dBzZyfZmuTmJM9f3u5IkvbFqN1Hb1/Ge3+EwZxJu553eE9VvXO4IckxwHrgWODRwBVJHr9zWg1J0niMej+FLwO3Agc3y9cA1+5lm68APx6xjpOBi6vqnqq6BdgKnDjitpKk/WTUqbNfA3wa+GDTdATw2WV+5uuSfLvpXnrk0PvdNvSahaZtsVrOSDKfZH7Hjh3LLEGStJhRzymcCZwE3AntDXcOW8bnfYDB0NbjgNuBdzXtWeS1i450qqoNVTVXVXOrV69eRgmSpN0ZNRTuqap7d64kOYjd/NHek6q6o6rur6pfMrg/w84uogUGt/rc6Uhg267bS5K6NWoofDnJOcCDm3szfwr416V+WJI1Q6svZzDUFeAyYH2SQ5IcDawDrl7q+0uS9s2oo4/OAk4HrgdeC3we+PCeNkjyCeCZwKokC8DbgGcmOY7BUcatzXtRVZuTbARuBO4DznTkkSSNX6b5IuW5ubman59f9vZZ7EyGJE2BffnTnWRTVc0t9tyocx/dwiLnEKrqscsvS5I0aZYy99FODwJOAR61/8uRJPVp1IvXfjT09YOqei/w7G5LkySN26jdRycMrR7A4MjhYZ1UJEnqzajdR+8aWr6PwcihV+z3aiRJvRr1dpzP6roQSVL/Ru0+etOenq+qd++fciRJfVrK6KOnM7jyGOAlwFf49UnsJElTbtRQWAWcUFV3weBmOcCnquovuipMkjR+o8599Bjg3qH1e4G1+70aSVKvRj1S+BhwdZJLGVzZ/HIeeEc1SdKUG3X00blJ/h34g6bp1VX1ze7KkiT1YdTuI4CHAHdW1T8AC80U15KkFWTU23G+DXgLcHbTdDDw8a6KkiT1Y9QjhZcDLwV+BlBV23CaC0lacUYNhXtrcOOFAkhyaHclSZL6MmoobEzyQeARSV4DXMHgHsuSpBVkr6OPkgT4JPBE4E7gCcDfVdXlHdcmSRqzvYZCVVWSz1bV0wCDQJJWsFG7j76e5OmdViJJ6t2oVzQ/C/jLJLcyGIEUBgcRT+mqMEnS+O0xFJI8pqq+D7xwTPVIknq0tyOFzzKYHfV7SS6pqj8dQ02SpJ7s7ZxChpYf22UhkqT+7S0UajfLkqQVaG/dR09NcieDI4YHN8vwqxPND++0OknSWO0xFKrqwHEVIknq31KmzpYkrXCGgiSp1VkoJDk/yfYkNwy1PSrJ5Um+0zw+cui5s5NsTXJzkud3VZckafe6PFL4CPCCXdrOAq6sqnXAlc06SY4B1gPHNtu8P4nnMyRpzDoLhar6CvDjXZpPBi5oli8AXjbUfnFV3VNVtwBbgRO7qk2StLhxn1M4vKpuB2geD2vajwBuG3rdQtP2AEnOSDKfZH7Hjh2dFitJs2ZSTjRnkbZFL5arqg1VNVdVc6tXr+64LEmaLeMOhTuSrAFoHrc37QvAUUOvOxLYNubaJGnmjTsULgNOa5ZPAz431L4+ySFJjgbWAVePuTZJmnmj3k9hyZJ8AngmsCrJAvA24B0M7vd8OvB94BSAqtqcZCNwI3AfcGZV3d9VbZKkxXUWClV16m6ees5uXn8ucG5X9UiS9m5STjRLkiaAoSBJahkKkqSWoSBJahkKkqSWoSBJahkKkqSWoSBJahkKkqSWoSBJahkKkqSWoSBJahkKkqSWoSBJahkKkqSWoSBJahkKkqSWoSBJahkKkqSWoSBJahkKkqSWoSBJahkKkqSWoSBJahkKkqSWoSBJahkKkqSWoSBJah3Ux4cmuRW4C7gfuK+q5pI8CvgksBa4FXhFVf2kj/okaVb1eaTwrKo6rqrmmvWzgCurah1wZbMuSRqjSeo+Ohm4oFm+AHhZf6VI0mzqKxQK+FKSTUnOaNoOr6rbAZrHwxbbMMkZSeaTzO/YsWNM5UrSbOjlnAJwUlVtS3IYcHmSm0bdsKo2ABsA5ubmqqsCJWkW9XKkUFXbmsftwKXAicAdSdYANI/b+6hNkmbZ2EMhyaFJHrZzGXgecANwGXBa87LTgM+NuzZJmnV9dB8dDlyaZOfnX1RVX0hyDbAxyenA94FTeqhNkmba2EOhqr4LPHWR9h8Bzxl3PZKkX5mkIamSpJ4ZCpKklqEgSWoZCpKklqEgSWoZCpKklqEgSWoZCpKklqEgSWoZCpKklqEgSWoZCpKklqEgSWoZCpKklqEgSWoZCpKklqEgSWoZCpKklqEgSWoZCpKklqEgSWoZCpKklqEgSWoZCpKklqEgSWoZCpKklqEgSWoZCpKklqEgSWpNXCgkeUGSm5NsTXJW3/VI0iyZqFBIciDwPuCFwDHAqUmO6bcqSZodExUKwInA1qr6blXdC1wMnNxzTZI0Mw7qu4BdHAHcNrS+APz28AuSnAGc0azeneTmffi8VcAP92H7SbAS9gHcj0mzEvZjJewD7GY/kn16z9/a3ROTFgqL7Wb92krVBmDDfvmwZL6q5vbHe/VlJewDuB+TZiXsx0rYBxj/fkxa99ECcNTQ+pHAtp5qkaSZM2mhcA2wLsnRSX4DWA9c1nNNkjQzJqr7qKruS/I64IvAgcD5VbW5w4/cL91QPVsJ+wDux6RZCfuxEvYBxrwfqaq9v0qSNBMmrftIktQjQ0GS1JrJUFgJU2kkOT/J9iQ39F3LvkhyVJKrkmxJsjnJG/quaamSPCjJ1Um+1ezD3/dd075IcmCSbyb5t75rWa4ktya5Psl1Seb7rme5kjwiyaeT3NT8jvxu5585a+cUmqk0/gd4LoMhsNcAp1bVjb0WtkRJngHcDXy0qp7cdz3LlWQNsKaqrk3yMGAT8LJp+n4kCXBoVd2d5GDga8AbqurrPZe2LEneBMwBD6+qF/ddz3IkuRWYq6qpvngtyQXAV6vqw82IzIdU1U+7/MxZPFJYEVNpVNVXgB/3Xce+qqrbq+raZvkuYAuDK9unRg3c3awe3HxN5X9bSY4E/hj4cN+1zLokDweeAZwHUFX3dh0IMJuhsNhUGlP1R2ilSrIWOB74Rs+lLFnT5XIdsB24vKqmbh8a7wX+Fvhlz3XsqwK+lGRTMzXONHossAP4l6Y778NJDu36Q2cxFPY6lYbGL8lDgUuAN1bVnX3Xs1RVdX9VHcfgKvwTk0xdl16SFwPbq2pT37XsBydV1QkMZlw+s+lunTYHAScAH6iq44GfAZ2fA53FUHAqjQnT9MNfAlxYVZ/pu5590Rze/yfwgn4rWZaTgJc2/fEXA89O8vF+S1qeqtrWPG4HLmXQbTxtFoCFoaPOTzMIiU7NYig4lcYEaU7Sngdsqap3913PciRZneQRzfKDgT8Cbuq1qGWoqrOr6siqWsvg9+I/qurPey5ryZIc2gxaoOlueR4wdaP0qup/gduSPKFpeg7Q+QCMiZrmYhx6mEqjE0k+ATwTWJVkAXhbVZ3Xb1XLchLwKuD6pk8e4Jyq+nx/JS3ZGuCCZmTbAcDGqpra4ZwrwOHApYP/NzgIuKiqvtBvScv2euDC5h/Y7wKv7voDZ25IqiRp92ax+0iStBuGgiSpZShIklqGgiSpZShI0pRY6kSYSV6R5MZmosaLRtrG0UeSNB2WMhFmknXARuDZVfWTJIc1F/PtkUcKkjQlFpsIM8njknyhmefpq0me2Dz1GuB9VfWTZtu9BgIYCpI07TYAr6+qpwF/Dby/aX888Pgk/5Xk60lGmnpl5q5olqSVoplI8veATzVXcAMc0jweBKxjMPPBkcBXkzx5b9NvGwqSNL0OAH7azNC7qwXg61X1C+CWJDczCIlr9vaGkqQp1Ewzf0uSU2AwwWSSpzZPfxZ4VtO+ikF30nf39p6GgiRNiWYizP8GnpBkIcnpwCuB05N8C9jMr+4k+UXgR0luBK4C/qaqfrTXz3BIqiRpJ48UJEktQ0GS1DIUJEktQ0GS1DIUJEktQ0GS1DIUJEmt/wcib+o7CQxI2AAAAABJRU5ErkJggg==",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     }
    }
   ],
   "metadata": {}
  },
  {
   "cell_type": "code",
   "execution_count": 19,
   "source": [
    "from IPython import display\r\n",
    "display.Image(\"boxplot.png\")"
   ],
   "outputs": [
    {
     "output_type": "execute_result",
     "data": {
      "image/png": "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",
      "text/plain": [
       "<IPython.core.display.Image object>"
      ]
     },
     "metadata": {},
     "execution_count": 19
    }
   ],
   "metadata": {}
  },
  {
   "cell_type": "code",
   "execution_count": 21,
   "source": [
    "gff['start'].plot.box()"
   ],
   "outputs": [
    {
     "output_type": "execute_result",
     "data": {
      "text/plain": [
       "<AxesSubplot:>"
      ]
     },
     "metadata": {},
     "execution_count": 21
    },
    {
     "output_type": "display_data",
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     }
    }
   ],
   "metadata": {}
  },
  {
   "cell_type": "markdown",
   "source": [
    "##  You can use Pandas library to generate basic statistical insights from data. For example, you can compute mean, standard deviation, correlation, minimum and maximum values, and so on.\r\n",
    "## Here, the mean of the starting positions is computed."
   ],
   "metadata": {}
  },
  {
   "cell_type": "code",
   "execution_count": 22,
   "source": [
    "gff['start'].mean()"
   ],
   "outputs": [
    {
     "output_type": "execute_result",
     "data": {
      "text/plain": [
       "3023697.813734794"
      ]
     },
     "metadata": {},
     "execution_count": 22
    }
   ],
   "metadata": {}
  },
  {
   "cell_type": "markdown",
   "source": [
    "## Find maximum and minimum of starting positions"
   ],
   "metadata": {}
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "source": [],
   "outputs": [],
   "metadata": {}
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "source": [],
   "outputs": [],
   "metadata": {}
  },
  {
   "cell_type": "markdown",
   "source": [
    "## What is the median of the starting and ending positions?"
   ],
   "metadata": {}
  },
  {
   "cell_type": "code",
   "execution_count": 38,
   "source": [
    "gff[['start','end']].median()"
   ],
   "outputs": [
    {
     "output_type": "execute_result",
     "data": {
      "text/plain": [
       "start    2989232.0\n",
       "end      2991801.0\n",
       "dtype: float64"
      ]
     },
     "metadata": {},
     "execution_count": 38
    }
   ],
   "metadata": {}
  },
  {
   "cell_type": "markdown",
   "source": [
    "## Some basic statistics are shown here. "
   ],
   "metadata": {}
  },
  {
   "cell_type": "code",
   "execution_count": 23,
   "source": [
    "gff['start'].describe()"
   ],
   "outputs": [
    {
     "output_type": "execute_result",
     "data": {
      "text/plain": [
       "count    1.159100e+04\n",
       "mean     3.023698e+06\n",
       "std      1.772444e+06\n",
       "min      1.000000e+00\n",
       "25%      1.488480e+06\n",
       "50%      2.989232e+06\n",
       "75%      4.574117e+06\n",
       "max      6.113694e+06\n",
       "Name: start, dtype: float64"
      ]
     },
     "metadata": {},
     "execution_count": 23
    }
   ],
   "metadata": {}
  },
  {
   "cell_type": "markdown",
   "source": [
    "## value_counts() function returns object containing counts of unique values.The number of genes, CDSs, tRNAs, and so on can be seen here.\r\n"
   ],
   "metadata": {}
  },
  {
   "cell_type": "code",
   "execution_count": 24,
   "source": [
    "gff['feature'].value_counts()"
   ],
   "outputs": [
    {
     "output_type": "execute_result",
     "data": {
      "text/plain": [
       "gene            5761\n",
       "CDS             5700\n",
       "exon              61\n",
       "tRNA              53\n",
       "binding_site       6\n",
       "ncRNA              4\n",
       "rRNA               3\n",
       "tmRNA              1\n",
       "region             1\n",
       "STS                1\n",
       "Name: feature, dtype: int64"
      ]
     },
     "metadata": {},
     "execution_count": 24
    }
   ],
   "metadata": {}
  },
  {
   "cell_type": "markdown",
   "source": [
    "## DataFrames are iterable, as are their columns or rows. Below we iterate over the elements in the 'attribute' column of the rows where feature=='gene'. The iteration is done using a list-comprehension. We can do something which each selected element as it is created during the for-loop that is happening within the list-comprehension. Here, because the selected elements are strings, we perform a string operation, .split(), to split the string every time a semi-colon appears. The result is a list of lists of the items in the string that were split.\r\n",
    "\r\n",
    "##  We operated on a DataFrame column full of strings and got in return a list of lists of strings! This is why it is important to understand the object types in Python, since it allows you to understand how and why this was possible."
   ],
   "metadata": {}
  },
  {
   "cell_type": "code",
   "execution_count": 25,
   "source": [
    "# split the string objects of those attribute rows\r\n",
    "[i.split(\";\") for i in gff[gff.feature == \"gene\"][\"attribute\"].head()]"
   ],
   "outputs": [
    {
     "output_type": "execute_result",
     "data": {
      "text/plain": [
       "[['ID=gene0',\n",
       "  'Name=HMPREF0004_0003',\n",
       "  'gbkey=Gene',\n",
       "  'locus_tag=HMPREF0004_0003',\n",
       "  'partial=true'],\n",
       " ['ID=gene1',\n",
       "  'Name=HMPREF0004_0004',\n",
       "  'gbkey=Gene',\n",
       "  'locus_tag=HMPREF0004_0004'],\n",
       " ['ID=gene2',\n",
       "  'Name=HMPREF0004_0005',\n",
       "  'gbkey=Gene',\n",
       "  'locus_tag=HMPREF0004_0005',\n",
       "  'partial=true'],\n",
       " ['ID=gene3',\n",
       "  'Name=HMPREF0004_0006',\n",
       "  'gbkey=Gene',\n",
       "  'locus_tag=HMPREF0004_0006'],\n",
       " ['ID=gene4',\n",
       "  'Name=HMPREF0004_0007',\n",
       "  'gbkey=Gene',\n",
       "  'locus_tag=HMPREF0004_0007']]"
      ]
     },
     "metadata": {},
     "execution_count": 25
    }
   ],
   "metadata": {}
  },
  {
   "cell_type": "markdown",
   "source": [
    "## The functions of pandas are not limited to those I've listed. If you're keen to know, you can learn more by visiting this  [website](https://pandas.pydata.org/pandas-docs/stable/reference/general_functions.html#)  or reading Chapter 5 of the book (Python for Data Analysis-DATA WRANGLING WITH PANDAS, NUMPY, AND IPYTHON-, Wes McKinney ) I'll send you.\r\n"
   ],
   "metadata": {}
  },
  {
   "cell_type": "markdown",
   "source": [
    "# gffpandas library"
   ],
   "metadata": {}
  },
  {
   "cell_type": "markdown",
   "source": [
    "## The Python library gffpandas ([gffpandas tutorial]( https://gffpandas.readthedocs.io/en/latest/tutorial.html#)  ) facilitates the work with GFF3 files. Thereby, different conditions can be choosen to process the annotation data, as e.g. to retain only the entries of a specific feature. The big advantages are that several functions and thus options, can be combined and that a GFF3 file or even csv or tsv file can be returned. "
   ],
   "metadata": {}
  }
 ],
 "metadata": {
  "orig_nbformat": 4,
  "language_info": {
   "name": "python",
   "version": "3.8.8",
   "mimetype": "text/x-python",
   "codemirror_mode": {
    "name": "ipython",
    "version": 3
   },
   "pygments_lexer": "ipython3",
   "nbconvert_exporter": "python",
   "file_extension": ".py"
  },
  "kernelspec": {
   "name": "python3",
   "display_name": "Python 3.8.8 64-bit ('base': conda)"
  },
  "interpreter": {
   "hash": "c0f70214c0dd213f07f54ee5d6e0ea644bdbba35113c9bfe8aaa0d1db03ad5dd"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 2
}