1 #ifndef _MUSSA_CLASS_H_
2 #define _MUSSA_CLASS_H_
3 // This file is part of the Mussa source distribution.
4 // http://mussa.caltech.edu/
5 // Contact author: Tristan De Buysscher, tristan@caltech.edu
7 // This program and all associated source code files are Copyright (C) 2005
8 // the California Institute of Technology, Pasadena, CA, 91125 USA. It is
9 // under the GNU Public License; please see the included LICENSE.txt
10 // file for more information, or contact Tristan directly.
13 // ----------------------------------------
14 // ---------- mussa_class.hh -----------
15 // ----------------------------------------
19 #include <boost/filesystem/path.hpp>
20 #include <boost/shared_ptr.hpp>
28 #include "alg/annotation_colors.hpp"
29 #include "alg/mussa_callback.hpp"
30 #include "alg/nway_paths.hpp"
31 #include "alg/sequence.hpp"
33 std::string int_to_str(int an_int);
36 //! provide a simple name to point to our Mussa shared_ptr
37 typedef boost::shared_ptr<Mussa> MussaRef;
39 class Mussa : public QObject
44 //! call whatever signaling system we want
45 void progress(const QString& description, int cur, int max);
46 //! triggered when our state changes between unsaved(true) and saved(false)
47 void isModified(bool);
50 typedef std::set<Sequence> motif_set;
51 enum analysis_modes { TransitiveNway, RadialNway, EntropyNway,
57 //! dynamically construct a new Mussa object and return a reference to it
58 static MussaRef init();
61 void save(boost::filesystem::path save_path="");
62 //! save the nway comparison
63 void save_muway(boost::filesystem::path save_path);
64 //! load a saved analysis directory
65 void load(boost::filesystem::path ana_path);
66 // ! return path to the where the analysis is stored
67 boost::filesystem::path get_analysis_path() const;
69 void set_analysis_path(boost::filesystem::path);
71 //! clear parameters and initialize data lists
74 //! set parameters from a file - 'mupa' ~ mussa parameters
75 void load_mupa_file(std::string para_file_path) { load_mupa_file(boost::filesystem::path(para_file_path));}
76 void load_mupa_file(boost::filesystem::path para_file_path);
78 // set parameters individually (eg from user input into gui classes)
80 void set_name(std::string a_name);
81 //! return name for this analysis
82 std::string get_name() const;
83 //! return a reasonable window title for this analysis
84 /*! this returns the "variable" portion for a title
86 std::string get_title() const;
88 //! return number of sequences in this analyzis
89 /*! this returns either the_seqs.size() or seq_files.size()
90 * depending on which has data loaded in
91 * (silly delayed loading of sequence data)
95 //! set number of bases for this window size
96 void set_window(int a_window);
97 //! get number of bases for the sliding window
98 int get_window() const;
99 //! set number of bases that must match for a window to be saved
100 //! if threshold > soft_threshold this also sets soft_threshold
101 void set_threshold(int a_threshold);
102 //! get number of bases that must match for a window to be saved
103 int get_threshold() const;
104 //! sets the threshold used for computing the nway paths
105 //! must be in range [threshold..window size]
106 void set_soft_threshold(int sft_thres);
107 int get_soft_threshold() const;
109 void set_analysis_mode(enum analysis_modes new_ana_mode);
110 enum analysis_modes get_analysis_mode() const;
111 //! return a string name for an analysis mode
112 std::string get_analysis_mode_name() const;
113 //! set our current dirty state
114 void set_dirty(bool);
115 //! return if we have unsaved changes
116 bool is_dirty() const;
117 //! is there anything loaded into this analysis?
120 //! return the refined paths found by the nway analysis.
121 const NwayPaths& paths() const;
123 //! given selected_paths, and view_paths, compute per base pair matches
124 //template <class IteratorT>
125 void createLocalAlignment(std::list<ConservedPath>::iterator begin,
126 std::list<ConservedPath>::iterator end,
127 std::list<ConservedPath::path_type>& result,
128 std::list<std::vector<bool> >& reversed);
130 //! run seqcomp and the nway filtering algorithm.
131 /*!analyze will run seqcomp and then the nway algorithm
132 * on whatever sequences have been loaded into this mussa instance.
133 * \throws mussa_analysis_error
136 /*! Run the nway filtering algorithm,
137 * this might be used when changing the soft threshhold?
141 //! appends a string sequence to the list of the_seqs
142 // void append_sequence(std::string a_seq);
143 //! appends a sequence to the list of the_seqs (makes copy)
144 void append_sequence(const Sequence& a_seq);
145 //! append a sequence to the list of seqs (shared)
146 void append_sequence(boost::shared_ptr<Sequence> a_seq);
148 //! Load a sequence from a fasta file and any annotations
149 /*! \param[in] seq_file the full path to the fasta file
150 * \param[in] annot_file the full path to an annotation file,
151 * if is an empty string, we won't bother loading anything
152 * \param[in] fasta_index specify which sequence in a multisequence fasta
154 * \param[in] sub_seq_start starting slice index to select a subsequence
155 * use 0 start from the beginning.
156 * \param[in] sub_seq_end ending slice index to select a subsequence
157 * use 0 to go to the end.
158 * \param[in] name sequence name, only used if not null
160 void load_sequence(boost::filesystem::path seq_file,
161 boost::filesystem::path annot_file,
162 int fasta_index, int sub_seq_start=0, int sub_seq_end=0,
163 std::string *name=0);
164 //! allow examining the sequences we have loaded
165 typedef std::vector<boost::shared_ptr<Sequence> > vector_sequence_type;
166 const vector_sequence_type& sequences() const;
168 // deprecated - support bridge for python version of mussa
169 // these save & load from the old file format
171 void load_old(char * load_file_path, int s_num);
173 // manage motif lists
174 //! add a motif it wont be applied until update_sequences_motif is called
175 void add_motif(const Sequence& motifs, const Color& colors);
176 //! add vector of motifs and colors to our motif collection
177 /*! this will automatically call update_sequences_motif
178 * this depends on sets and color maps being unique
179 * (aka if you add the same item more than once it doesn't
180 * increase the size of the data structure)
182 void set_motifs(const std::vector<Sequence>& motifs,
183 const std::vector<Color>& colors);
184 /*! The file should look something like
185 * <sequence> <red> <green> <blue>
186 * where sequence is a string of IUPAC symbols
187 * and red,green,blue are a white space separated list of floats
188 * in the range [0.0, 1.0]
190 //! load a list of motifs from a file named filename
191 void load_motifs(boost::filesystem::path filename);
192 //! load motifs from an ifstream
193 /*! \sa Mussa::load_motifs(boost::filesystem::path)
195 void load_motifs(std::istream &);
196 //! save motif list to the specified filename
197 void save_motifs(boost::filesystem::path filename);
198 //! save motif list to an ostream
199 void save_motifs(std::ostream &);
200 //! return our motifs;
201 const motif_set& motifs() const;
203 //! return color mapper
204 boost::shared_ptr<AnnotationColors> colorMapper();
207 //! push motifs to our attached sequences
208 void update_sequences_motifs();
211 // parameters needed for a mussa analysis
212 //! name of this analysis. (will also be used when saving an analysis)
213 std::string analysis_name;
214 //! how many base pairs to include in a sliding window
216 //! how many base pairs need to match order to record a window as conserved
218 //! stores current filter used by GUI to change the connections shown
220 //! which nway comparison algorithm to use.
221 enum analysis_modes ana_mode;
223 //! should we append _w<window_size> to the saved analysis
225 //! should we append _t<threshold> to the saved analysis
229 vector_sequence_type the_seqs;
231 std::vector<std::vector<FLPs> > all_comps;
232 //! N-way data, ie the mussa results
236 motif_set motif_sequences;
238 boost::shared_ptr<AnnotationColors> color_mapper;
239 //! path to our analysis
240 boost::filesystem::path analysis_path;
241 //! flag indicating if we have unsaved changes
245 //! runs all the seqcomps needed to support the nway comparison