Demo - How to run Variant calling pipeline?.ipynb

{
 "cells": [
  {
   "cell_type": "code",
   "execution_count": 12,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div style=\"text-align:center;background:#9999ff;padding:60px;color:#ffffff;font-size:2em;\"></div>"
      ],
      "text/plain": [
       "<IPython.core.display.HTML object>"
      ]
     },
     "execution_count": 12,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "from IPython.display import HTML\n",
    "def new_section(title):\n",
    "    style = \"text-align:center;background:#9999ff;padding:60px;color:#ffffff;font-size:2em;\"\n",
    "    return HTML('<div style=\"{}\">{}</div>'.format(style, title))\n",
    "new_section(\"\")"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "This is a quick demo of how to run variant calling pipeline on a test dataset. The raw test dataset for this exercise can be found in:\n",
    "\n",
    "\n",
    "```\n",
    "/nfs/esnitkin/bin_group/variant_calling_bin/test_data\n",
    "```"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 13,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "/nfs/esnitkin/bin_group/variant_calling_bin/test_data/MRSA_jail_100_R1.fastq.gz\r\n",
      "/nfs/esnitkin/bin_group/variant_calling_bin/test_data/MRSA_jail_10_R1.fastq.gz\r\n",
      "/nfs/esnitkin/bin_group/variant_calling_bin/test_data/MRSA_jail_1_R1.fastq.gz\r\n"
     ]
    }
   ],
   "source": [
    "% ls -1a /nfs/esnitkin/bin_group/variant_calling_bin/test_data/*_R1.fastq.gz"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Step 1: \n",
    "\n",
    "We will call variants on all sets of PE reads in test_data folder. The script will detect the file extension of fastq file, generate individual job for each sample in test_data folder and submit these cluster jobs. Make sure to provide upur own config file that contains your user specific PBS values and tool parameters."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 12,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "START: Variant Calling Pipeline\n",
      "Generating cluster jobs in temporary directory /nfs/esnitkin/bin_group/variant_calling_bin/test_output//temp_jobs\n",
      "Running Jobs in parallel-cluster mode\n",
      "Submitting Job: qsub /nfs/esnitkin/bin_group/variant_calling_bin/test_output//temp_jobs/MRSA_jail_100.pbs\n",
      "Submitting Job: qsub /nfs/esnitkin/bin_group/variant_calling_bin/test_output//temp_jobs/MRSA_jail_10.pbs\n",
      "Submitting Job: qsub /nfs/esnitkin/bin_group/variant_calling_bin/test_output//temp_jobs/MRSA_jail_1.pbs\n",
      "You can check the job status with: qstat -u USERNAME\n",
      "Logs were recorded in file with extension log.txt in /nfs/esnitkin/bin_group/variant_calling_bin/test_output//Logs/variant_calling\n",
      "Total Time taken: 0:00:00.271331\n",
      "End: Variant calling Pipeline\n"
     ]
    }
   ],
   "source": [
    "! /nfs/esnitkin/bin_group/anaconda2/bin//python /nfs/esnitkin/bin_group/pipeline/Github/variant_calling_pipeline/variant_call.py -type PE -readsdir /nfs/esnitkin/bin_group/variant_calling_bin/test_data/ -outdir /nfs/esnitkin/bin_group/variant_calling_bin/test_output -analysis test_results -index MRSA_USA_300 -steps All -cluster parallel-cluster -config /nfs/esnitkin/bin_group/pipeline/Github/variant_calling_pipeline/config_ali"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Step 2: \n",
    "\n",
    "When all the submitted jobs are completed, you can move on to the next step of the analysis i.e core_prep. Run the same command with core_prep as your -steps argument"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 1,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "START: Extract core snps and generate diagnostic plots\n",
      "Copying vcf files to /nfs/esnitkin/bin_group/variant_calling_bin/test_output//core_temp_dir/\n",
      "Decompressing gzipped files in /nfs/esnitkin/bin_group/variant_calling_bin/test_output//core_temp_dir/\n",
      "Running command: gzip -df /nfs/esnitkin/bin_group/variant_calling_bin/test_output//core_temp_dir/MRSA_jail_100_aln_mpileup_raw.vcf_5bp_indel_removed.vcf.gz\n",
      "Running command: gzip -df /nfs/esnitkin/bin_group/variant_calling_bin/test_output//core_temp_dir/MRSA_jail_10_aln_mpileup_raw.vcf_5bp_indel_removed.vcf.gz\n",
      "Running command: gzip -df /nfs/esnitkin/bin_group/variant_calling_bin/test_output//core_temp_dir/MRSA_jail_1_aln_mpileup_raw.vcf_5bp_indel_removed.vcf.gz\n",
      "Running command: gzip -df /nfs/esnitkin/bin_group/variant_calling_bin/test_output//core_temp_dir/MRSA_jail_100_filter2_final.vcf.gz\n",
      "Running command: gzip -df /nfs/esnitkin/bin_group/variant_calling_bin/test_output//core_temp_dir/MRSA_jail_10_filter2_final.vcf.gz\n",
      "Running command: gzip -df /nfs/esnitkin/bin_group/variant_calling_bin/test_output//core_temp_dir/MRSA_jail_1_filter2_final.vcf.gz\n",
      "Running command: gzip -df /nfs/esnitkin/bin_group/variant_calling_bin/test_output//core_temp_dir/MRSA_jail_100_filter2_final.vcf_no_proximate_snp.vcf.gz\n",
      "Running command: gzip -df /nfs/esnitkin/bin_group/variant_calling_bin/test_output//core_temp_dir/MRSA_jail_10_filter2_final.vcf_no_proximate_snp.vcf.gz\n",
      "Running command: gzip -df /nfs/esnitkin/bin_group/variant_calling_bin/test_output//core_temp_dir/MRSA_jail_1_filter2_final.vcf_no_proximate_snp.vcf.gz\n",
      "Submitting parallel-cluster Job: qsub /nfs/esnitkin/bin_group/variant_calling_bin/test_output//core_temp_dir//2018_04_26_11_17_50_test_results.pbs\n",
      "28934701\n",
      "You can check the job status with: qstat -u USERNAME\n",
      "Logs were recorded in file with extension log.txt in /nfs/esnitkin/bin_group/variant_calling_bin/test_output//Logs/core_prep\n",
      "Total Time taken: 0:00:02.039515\n"
     ]
    }
   ],
   "source": [
    "! /nfs/esnitkin/bin_group/anaconda2/bin//python /nfs/esnitkin/bin_group/pipeline/Github/variant_calling_pipeline/variant_call.py -type PE -readsdir /nfs/esnitkin/bin_group/variant_calling_bin/test_data/ -outdir /nfs/esnitkin/bin_group/variant_calling_bin/test_output -analysis test_results -index MRSA_USA_300 -steps core_prep -cluster parallel-cluster -config /nfs/esnitkin/bin_group/pipeline/Github/variant_calling_pipeline/config_ali"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Step 3:\n",
    "\n",
    "This step will take all the intermediate files generated during the core_prep step and generate the core SNPs and various diagnostics plot related data matrices."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "START: Extract core snps and generate diagnostic plots\n",
      "Logs were recorded in file with extension log.txt in /nfs/esnitkin/bin_group/variant_calling_bin/test_output//Logs/core\n",
      "Total Time taken: 0:12:02.726169\n"
     ]
    }
   ],
   "source": [
    "! /nfs/esnitkin/bin_group/anaconda2/bin//python /nfs/esnitkin/bin_group/pipeline/Github/variant_calling_pipeline/variant_call.py -type PE -readsdir /nfs/esnitkin/bin_group/variant_calling_bin/test_data/ -outdir /nfs/esnitkin/bin_group/variant_calling_bin/test_output -analysis test_results -index MRSA_USA_300 -steps core -cluster parallel-cluster -config /nfs/esnitkin/bin_group/pipeline/Github/variant_calling_pipeline/config_ali"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Step 4:\n",
    "    \n",
    "Generate the pipeline report and transfer the final results to a *_core_results folder arranged in a more organized manner.  "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "START: Generating alignment and variant calling report\n",
      "/nfs/esnitkin/bin_group/variant_calling_bin/test_output//2018_04_26_11_23_59_core_results\n",
      "Logs were recorded in file with extension log.txt in /nfs/esnitkin/bin_group/variant_calling_bin/test_output//Logs/report\n",
      "Total Time taken: 0:00:04.932519\n"
     ]
    }
   ],
   "source": [
    "! /nfs/esnitkin/bin_group/anaconda2/bin//python /nfs/esnitkin/bin_group/pipeline/Github/variant_calling_pipeline/variant_call.py -type PE -readsdir /nfs/esnitkin/bin_group/variant_calling_bin/test_data/ -outdir /nfs/esnitkin/bin_group/variant_calling_bin/test_output -analysis test_results -index MRSA_USA_300 -steps report -cluster parallel-local -config /nfs/esnitkin/bin_group/pipeline/Github/variant_calling_pipeline/config_ali"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "All the final results and reports will be tranferred and arranged in:\n",
    "\n",
    "/nfs/esnitkin/bin_group/variant_calling_bin/test_output//2018_04_26_11_23_59_core_results"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      ".\r\n",
      "..\r\n",
      "core_snp_consensus\r\n",
      "data_matrix\r\n"
     ]
    }
   ],
   "source": [
    "! ls -1a /nfs/esnitkin/bin_group/variant_calling_bin/test_output/2018_04_26_11_23_59_core_results/"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 6,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      ".\r\n",
      "..\r\n",
      "consensus_ref_variant_positions\r\n",
      "consensus_variant_positions\r\n",
      "MRSA_jail_100_filter2_final.vcf_core.vcf.gz\r\n",
      "MRSA_jail_10_filter2_final.vcf_core.vcf.gz\r\n",
      "MRSA_jail_1_filter2_final.vcf_core.vcf.gz\r\n"
     ]
    }
   ],
   "source": [
    "! ls -1a /nfs/esnitkin/bin_group/variant_calling_bin/test_output/2018_04_26_11_23_59_core_results/core_snp_consensus/"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "core_snp_consensus folder will contain core variant vcf files and consesnsus variants in fasta format. consensus_ref_variant_positions will contain core variants as well as reference allele which can be used as an input for Gubbins. \n",
    "consensus_variant_positions will contain only core variants in fasta format that can be used for Beast and Fasttree analysis."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 7,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      ".\r\n",
      "..\r\n",
      "MRSA_jail_100.fa\r\n",
      "MRSA_jail_10.fa\r\n",
      "MRSA_jail_1.fa\r\n"
     ]
    }
   ],
   "source": [
    "! ls -1a /nfs/esnitkin/bin_group/variant_calling_bin/test_output/2018_04_26_11_23_59_core_results/core_snp_consensus/consensus_ref_variant_positions"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 8,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      ".\r\n",
      "..\r\n",
      "MRSA_jail_100_variants.fa\r\n",
      "MRSA_jail_10_variants.fa\r\n",
      "MRSA_jail_1_variants.fa\r\n",
      "MRSA_USA_300_variants.fa\r\n"
     ]
    }
   ],
   "source": [
    "! ls -1a /nfs/esnitkin/bin_group/variant_calling_bin/test_output/2018_04_26_11_23_59_core_results/core_snp_consensus/consensus_variant_positions"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "As the name suggests, data_matrix contains all the matrices that were used for generating the core variants and which can be used for plotting various diagnostics plots."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 10,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      ".\r\n",
      "..\r\n",
      "2018_04_26_11_22_59_step_1.log.txt\r\n",
      "2018_04_26_11_24_09_step_2.log.txt\r\n",
      "2018_04_26_11_47_10_step_3.log.txt\r\n",
      "matrices\r\n",
      "Reports\r\n",
      "snpeff_reports\r\n"
     ]
    }
   ],
   "source": [
    "! ls -1a /nfs/esnitkin/bin_group/variant_calling_bin/test_output/2018_04_26_11_23_59_core_results/data_matrix/"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "matrices contain all the diagnostic plot related data. Reports folder contain a concise variants report of variants found in each sample, alignments related information and SNP/Indel matrix. snpeff_reports contain annotation report generated by snpEFF."
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Lets plot some diagnostics plots that can use to troubleshoot variant calling pipeline.\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 11,
   "metadata": {},
   "outputs": [],
   "source": [
    "! module load R"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 17,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "/nfs/esnitkin/bin_group/variant_calling_bin/test_output/2018_04_26_11_23_59_core_results/data_matrix/matrices\n"
     ]
    }
   ],
   "source": [
    "%cd /nfs/esnitkin/bin_group/variant_calling_bin/test_output/2018_04_26_11_23_59_core_results/data_matrix/matrices/"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 44,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "null device \r\n",
      "          1 \r\n"
     ]
    }
   ],
   "source": [
    "! Rscript bargraph.R"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "This will generate a barplot.png file showing the distribution of each variant filters on the final variants."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 45,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<IPython.core.display.Image object>"
      ]
     },
     "execution_count": 45,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "from IPython.display import Image\n",
    "fig = Image(filename=('./barplot.png'))\n",
    "fig"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": []
  }
 ],
 "metadata": {
  "kernelspec": {
   "display_name": "Python 2",
   "language": "python",
   "name": "python2"
  },
  "language_info": {
   "codemirror_mode": {
    "name": "ipython",
    "version": 3
   },
   "file_extension": ".py",
   "mimetype": "text/x-python",
   "name": "python",
   "nbconvert_exporter": "python",
   "pygments_lexer": "ipython3",
   "version": "3.5.4"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 2
}