windowExtractor.h 3.72 KB
Newer Older
Mathieu Giraud's avatar
Mathieu Giraud committed
1 2 3 4
#ifndef WINDOW_EXTRACTOR_H
#define WINDOW_EXTRACTOR_H

#include <iostream>
5 6
#include <string>
#include <map>
Mathieu Giraud's avatar
Mathieu Giraud committed
7
#include "segment.h"
8
#include "germline.h"
Mathieu Giraud's avatar
Mathieu Giraud committed
9 10 11 12 13 14 15 16 17 18 19 20 21
#include "kmerstore.h"
#include "kmeraffect.h"
#include "windows.h"

using namespace std;

/**
 * This takes an OnlineFasta reader as input and extract windows from the
 * sequences given in the input.
 */
class WindowExtractor {
 private:
  size_t nb_reads;
22
  map<string, size_t> nb_reads_germline;
Mathieu Giraud's avatar
Mathieu Giraud committed
23 24 25

  ostream *out_segmented;
  ostream *out_unsegmented;
26
  ostream *out_affects;
Mathieu Giraud's avatar
Mathieu Giraud committed
27

28
  Stats stats[STATS_SIZE];
29
  size_t max_reads_per_window;
Mathieu Giraud's avatar
Mathieu Giraud committed
30 31 32 33 34 35 36 37 38
 public:

  WindowExtractor();

  /**
   * Extract windows from the collection of input reads.
   * If (un)segmented sequences must be output, the functions 
   * set(Un)SegmentedOutput() must be called before.
   * @param reads: the collection of input reads
39
   * @param multigermline: the multigermline
Mathieu Giraud's avatar
Mathieu Giraud committed
40 41
   * @param w: length of the window
   * @param windows_labels: Windows that must be kept and registered as such.
42
   * @param nb_expected: maximal e-value of the segmentation
43
   * @param nb_reads_for_evalue: number of reads, used for e-value computation. Can be approximate or faked.
Mathieu Giraud's avatar
Mathieu Giraud committed
44 45 46 47 48
   * @return a pointer to a WindowsStorage that will contain all the windows.
   *         It is a pointer so that the WindowsStorage is not duplicated.
   * @post Statistics on segmentation will be provided through the getSegmentationStats() methods
   *       and getAverageSegmentationLength().
   */
49 50
  WindowsStorage *extract(OnlineFasta *reads, MultiGermline *multigermline,
                          size_t w,
51
                          map<string, string> &windows_labels,
52
                          int stop_after=-1, int only_nth_reads=1, bool keep_unsegmented_as_clone=false,
53
                          double nb_expected = THRESHOLD_NB_EXPECTED, int nb_reads_for_evalue = 1);
Mathieu Giraud's avatar
Mathieu Giraud committed
54 55 56 57 58 59 60 61

  /**
   * @return the average length of sequences whose segmentation has been classified as seg
   * @param seg: one of the segmentation available in the enum SEGMENTED 
   * @pre extract() must have been launched.
   */
  float getAverageSegmentationLength(SEGMENTED seg);

62 63 64 65 66
  /**
   * cf. WindowsStorage::getMaximalNbReadsPerWindow()
   */
  size_t getMaximalNbReadsPerWindow();

Mathieu Giraud's avatar
Mathieu Giraud committed
67 68 69 70 71 72 73 74 75 76 77 78
  /**
   * @return Total number of processed reads by the previous call to extract()
   */
  size_t getNbReads();

  /**
   * @return the number of sequences whose segmentation has been classified as seg
   * @param seg: one of the segmentation available in the enum SEGMENTED 
   * @pre extract() must have been launched.
   */
  size_t getNbSegmented(SEGMENTED seg);

79 80 81 82 83 84
   /**
   * @return the number of reads segmented from germline
   * @param germline_code: one of the germline code in multigermline
   * @pre extract() must have been launched.
   */
  size_t getNbReadsGermline(string germline_code);
85 86 87 88 89 90

  /**
   * cf. WindowsStorage::setMaximalNbReadsPerWindow()
   */
  void setMaximalNbReadsPerWindow(size_t max_reads);

Mathieu Giraud's avatar
Mathieu Giraud committed
91 92 93 94 95 96 97 98 99 100 101 102 103 104
  /**
   * Defines the output stream where the segmented sequences will be output.
   * Otherwise no output will be given.
   * @param out: The output stream
   */
  void setSegmentedOutput(ostream *out);

  /**
   * Defines the output stream where the unsegmented sequences will be output.
   * Otherwise no output will be given.
   * @param out: The output stream
   */
  void setUnsegmentedOutput(ostream *out);

105 106 107 108 109 110 111
  /**
   * Defines the output stream where the detailed affects will be output.
   * Otherwise no output will be given.
   * @param out: The output stream
   */
  void setAffectsOutput(ostream *out);

112 113 114 115 116 117
  /**
   * Output the segmentation stats
   * @param out: The output stream
   */
  void out_stats(ostream &out);

Mathieu Giraud's avatar
Mathieu Giraud committed
118 119 120 121 122 123 124 125
 private:
  /**
   * Initialize the statistics (put 0 everywhere).
   */
  void init_stats();
};

#endif