]> git.donarmstrong.com Git - mothur.git/blob - chimera.h
working on chimeras
[mothur.git] / chimera.h
1 #ifndef CHIMERA_H
2 #define CHIMERA_H
3
4 /*
5  *  chimera.h
6  *  Mothur
7  *
8  *  Created by Sarah Westcott on 7/9/09.
9  *  Copyright 2009 Schloss Lab UMASS Amherst. All rights reserved.
10  *
11  */
12
13
14 #include "mothur.h"
15 #include "sparsematrix.hpp"
16 #include "sequence.hpp"
17
18 typedef list<PCell>::iterator MatData;
19 typedef map<int, float> SeqMap;  //maps sequence to all distance for that seqeunce
20
21
22 /***********************************************************************/
23
24 class Chimera {
25
26         public:
27         
28                 Chimera(){};
29                 Chimera(string);
30                 Chimera(string, string);
31                 virtual ~Chimera(){};
32                 virtual void setFilter(bool f)                  {       filter = f;                     }
33                 virtual void setCorrection(bool c)              {       correction = c;         }
34                 virtual void setProcessors(int p)               {       processors = p;         }
35                 virtual void setWindow(int w)                   {       window = w;                     }
36                 virtual void setIncrement(int i)                {       increment = i;          }
37                 
38                 virtual void setCons(string) {};
39                 virtual void setQuantiles(string) {};
40                 
41                 virtual vector<Sequence*> readSeqs(string file) {
42                         try {
43                                 ifstream in;
44                                 openInputFile(file, in);
45                                 vector<Sequence*> container;
46                                 
47                                 //read in seqs and store in vector
48                                 while(!in.eof()){
49
50                                         Sequence* current = new Sequence(in);
51                                         container.push_back(current);
52                                         gobble(in);
53                                 }
54                                 
55                                 in.close();
56                                 return container;
57                         }
58                         catch(exception& e) {
59                                 errorOut(e, "Chimera", "readSeqs");
60                                 exit(1);
61                         }
62                 }
63                 
64                 
65                 virtual void setMask(string filename) {
66                         try {
67                                 
68                                 if (filename == "default") {
69                                         //default is from wigeon  236627 EU009184.1 Shigella dysenteriae str. FBD013
70                                         seqMask = ".....................................................................................................AAATTGAAGAGTTT-GA--T-CA-T-G-GCTC-AG-AT-TGAA-C-GC--TGG-C--G-GC-A-GG--C----C-T--AACACA-T-GC-A-AGT-CGA-A-CG----------G-TAA-CA-G----------------------------GAAG-A-AG----------------------------------------------------CTT-G----------------------------------------------------------------------------------CT-TCTTT----------------G-CT--G--AC--G--AG-T-GG-C-GG-A--C-------------GGG-TGAGT-A--AT-GT-C-T-G-GG---A-A--A-CT-G--C-C-TGA--TG-G------------------------------------------------------------------A-GG----GGG-AT-AA-CTA-------------------------C-T-G-----------------------GAA-A---CGG-TAG-CTAA-TA---CC-G--C-AT-A----------A--------------------C-------------------------------------GT-C-----------------------------------------------------------------------------------------------------------------------G-CA-A--------------------------------------------------------------------------------------------------------------------------------------G-A-C---------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------CAAA--G-A-G-GG-----G--GA-C-CT--------------------------------------------------------------------------------------------------------------------TCG-G----------------------------------------------------------------------------------------------------------------------G----CC-TC--T---T-G--------------C----C-A---T-CG-G---AT---G-T-----G-CCC-AGA--T-GGG--A------TT--A--G-CT-A----G---TAGG-T-G-GG-G-T----AAC-GG-C-T-C-ACCT--A-GG-C-G--A-CG-A------------TCC-C-T------AG-CT-G-G-TCT-G-AG----A--GG-AT--G-AC-C-AG-CCAC-A-CTGGA--A-C-TG-A-GA-C-AC-G-G-TCCAGA-CTCC-TAC-G--G-G-A-G-GC-A-GC-A-G-TG---GG-G-A-ATA-TTGCA-C-AA-T-GG--GC-GC-A----A-G-CC-T-GA-TG-CA-GCCA-TGCC-G-CG-T---G-T-A--T--GA-A-G--A--A-G-G-CC-----TT-CG---------G-G-T-T-G-T--A---AA-G-TAC--------TT-TC-A-G--C-GGG----GA-G--G---AA-GGGA---GTAA-AG----T--T--AA-T---A----C-----CT-T-TGC-TCA-TT-GA-CG-TT-A-C-CC-G-CA-G---------AA-----------GAAGC-ACC-GG-C-TAA---C--T-CCGT--GCCA--G-C---A--GCCG---C-GG--TA-AT--AC---GG-AG-GGT-GCA-A-G-CG-TTAA-T-CGG-AA-TT-A--C-T--GGGC-GTA----AA-GCGC-AC--G-CA-G-G-C-G------------G--T-TT-G-T-T-AA----G-T-C-A---G-ATG-TG-A-AA-TC--CC-CGG-G--------------------------------------------------------------------CT-C-AA-------------------------------------------------------------------------CC-T-G-GG-AA-C----T-G-C-A-T-C--------T--GA-T-A-C-T-G-GCA--A-G-C---------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------T-T-G-A-G-T-C-----T-CG--TA-G-A------------G-GG-G-GG-T----AG--AATT-CCA-G-GT--GT-A-GCG-GTGAAA-TG-CGT-AGAG-A-TC-T-GGA--GG-A-AT-A-CC-GG--T--G--GC-GAA-G--G-C---G----G--C-C-CCCTG------G-AC-GA--------------------------------------------------------------AG-A-C-T--GA--CG-----CT-CA-GG--T-G-CGA--AA-G-C--------------G-TGGG-GAG-C-A-AACA--GG-ATTA-G-ATA-C-----CC-T-G-GTA-G-T----C-CA--C-G-CCG-T-AAA--C-GATG-TC--GA-CT---------T-GG--A--G-G-TT-G-TG-C--C--------------------------------------------------------------------------------------CTT-GA--------------------------------------------------------------------------------------------------------------------------------------------------G-G-C-GT--G-G-C-T-TC-C------GG--A----GC-TAA--CG-C-G-T--T--AA-GT--C----G-ACC-GCC-T-G-GG-GAG-TA---CGG-----C-C--G-C-A-A-GGT-T--AAA-ACTC-AAA---------TGAA-TTG-ACGGG-G-G-CCCG----C-A--C-A-A-GCG-GT-G--G--AG-CA-T--GT-GGT-TT-AATT-C-G-ATG-CAAC-G-CG-A-AG-A-A-CC-TT-A-CC-TGGTC-TT-G-AC-A-T-C--------------CAC-G-G-------------A-AG-T-T-T--TC--A-GA-G-A-T--G-A-G--A-A-T-G--T-G-----CC-------------------------------------T--TC-G------------------------------------------GG----A----A---CC-GTG---A--GA---------------------------------------------------C-A-G-G-T-GCTG-CA-TGG-CT--GTC-GTC-A-GC-TC---G-TG-TT-G--TGA-AA-TGT-T-GG-G-TT-AA-GT-CCCGC-AA--------C-GAG-CGC-A-ACC-C-T-TA--TC--C-TTTG--T-T-G-C-C---AG-C-G-----G-------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------TCC------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------GG---C----C-G------------G----G---A-A--CT---------------C-A-A-A-G-GA-G--AC-T-G-CCA--G-T------------------------------------G-A---TAA----------------------------------A-C-T-G--G-A-GG-A--AGG-T--GGGG-A-TGAC-GTC--AAGT-C---ATC-A-T-G-G-C-C-CTT----AC-G--AC-C-A-GG-GC-TA-CAC-ACGTG-C--TA--CAATG---G-CGCA-T-A--C-AAA-GA-GA--------------------------------------------------------------------------------------------------A-G-C-G-A--C-CTCG-C--G---------------------------------------A-GA-G-C-----------A--A-G-CG---G----------A--CCT-C------A-T-AAAGT-GC-G-T-C-G-TAG-TCC--------GGA-T-TGGAG-TC--T-GCAA-CT-C-------------------------------------------------------------------------------------------------G-ACTCC-A-T-G-AA-G-TC-GGAAT-CG-C-TA--G-TA-AT-C-G-T----GGA-TC-A-G--A------AT--GCC-AC-G-GT-G-AAT-ACGT-T-CCCGGGCCT-TGTA----CACACCG-CCC-GTC-----A---CA--CCA-TG-GG-A--G---TGG-G-TT-GC-AAA--A-GAA------G--T-AGG-TA-G-C-T-T-AA-C-C--------------------------------------------------------------TT----C-------------------------------------------------------------------------------------------------G--GG-A--GG-G--C---GC-TTA--CC--ACT-T----T-GTG-AT-TCA------------------------TG--ACT-GGGG-TG-AAG-TCGTAACAA-GGTAA-CCGT-AGGGGAA-CCTG-CGGT-TGGATCACCTCCTTA................................................................................................................................................................................................................................................................................................................................................................................................................................................................................................................................................................................................................................................................................................................................................................................................................................................................";
71                                 }else{
72                                         ifstream infile;
73                                         openInputFile(filename, infile);
74                                         
75                                         while (!infile.eof()) {
76                                                 Sequence temp(infile);
77                                                 seqMask = temp.getAligned();
78                                                 
79                                                 gobble(infile);
80                                         }
81                                         
82                                         infile.close();
83                                 }
84                         }
85                         catch(exception& e) {
86                                 errorOut(e, "Chimera", "setMask");
87                                 exit(1);
88                         }
89                 }
90                 
91                 //pure functions
92                 virtual void getChimeras() = 0; 
93                 virtual void print(ostream&) = 0;       
94                 
95         protected:
96                 
97                 bool filter, correction;
98                 int processors, window, increment;
99                 string seqMask;
100                         
101
102 };
103
104 /***********************************************************************/
105
106 #endif
107