{
 "cells": [
  {
   "cell_type": "code",
   "execution_count": 1,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "### parse sam file input\n",
    "### if too many reads unmapped\n",
    "\"\"\"\n",
    "Use only [] for indexing\n",
    "\"\"\"\n",
    "import sys\n",
    "import pandas as pd\n",
    "import os\n",
    "#fname='out.sam'\n",
    "fname=str(sys.argv[1])\n",
    "#specie='Homo_sapiens'#\n",
    "specie=str(sys.argv[2])\n",
    "emptyPickleDir='/cellar/users/btsui/Data/Project/Skymap/ChipSeq/empty_references/'+specie+'.pickle'\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "metadata": {
    "collapsed": false
   },
   "outputs": [],
   "source": [
    "myCountS=pd.read_pickle(emptyPickleDir)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "metadata": {
    "collapsed": false
   },
   "outputs": [],
   "source": [
    "\"\"\"\n",
    "\"\"\"\n",
    "##open sam file\n",
    "checkMark=100000\n",
    "with open(fname,'r') as f:\n",
    "    failedCount=0\n",
    "    for l in f:\n",
    "        if l[0]!='@':\n",
    "            break\n",
    "    for i,l in enumerate(f):\n",
    "        splitL=l.split('\\t')\n",
    "        start_loc=int(splitL[3])\n",
    "        tlen=len(splitL[9])\n",
    "        rname=splitL[2]\n",
    "        if rname !='*':\n",
    "            end_loc=start_loc+tlen\n",
    "            binned_start_loc=int(start_loc/20.0)*20\n",
    "            binned_end_loc=int(end_loc/20.0)*20\n",
    "            #print start_loc,end_loc\n",
    "            for i in range(binned_start_loc,binned_end_loc+1,20):\n",
    "                #print (rname,i)\n",
    "                myCountS[(rname,i)]+=1\n",
    "        else:\n",
    "            failedCount+=1\n",
    "            if i==checkMark:\n",
    "                frac=failedCount/float(checkMark)\n",
    "                if frac>0.9:\n",
    "                    break\n",
    "        #tlen=  int(int(splitL[9])/20.0)*20\n",
    "        #rname,pos,tlen=splitL[2],binnedLoc,splitL[9]\n",
    "            \n",
    "        "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 8,
   "metadata": {
    "collapsed": false
   },
   "outputs": [],
   "source": [
    "nonZero_CountS=myCountS[myCountS!=0]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 12,
   "metadata": {
    "collapsed": false
   },
   "outputs": [],
   "source": [
    "nonZero_CountS.to_pickle('bin_count.pickle')"
   ]
  }
 ],
 "metadata": {
  "kernelspec": {
   "display_name": "Python 2",
   "language": "python",
   "name": "python2"
  },
  "language_info": {
   "codemirror_mode": {
    "name": "ipython",
    "version": 2
   },
   "file_extension": ".py",
   "mimetype": "text/x-python",
   "name": "python",
   "nbconvert_exporter": "python",
   "pygments_lexer": "ipython2",
   "version": "2.7.13"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 0
}