#ifndef _DEF_PROFILE_H_
#define _DEF_PROFILE_H_

#include<stdio.h>
#include<dirent.h>
#include"seq_util.h"
#include"extreme_ml.h"

/* PROFILE_ALIGNMENT is used to describe the alignment of a sequence to a profile. Alignments are arrays of PROFILE_ALIGNMENT objects */

typedef struct {
  int pos1, pos2, type; /* pos1, pos2 = coords in sequence and profile. type is of BACKTRACK, giving path direction */
} 
PROFILE_ALIGNMENT;


typedef struct {
  int pos, score;
} WORD_HIT;

/* PREPROCESS is a struct containing linked lists of all word types of length 3 scoring more than threshold in a profile or sequence */

typedef struct {
  int *mapping;
  WORD_HIT **hitlist;
  int *hits;
  int total;
  int threshold;
  int diag_threshold;
  int max_dist;
  int window;
  struct profile *psi;
} PREPROCESS;

#define A1_SIZE 21
#define A2_SIZE 441
#define A3_SIZE 9261

typedef struct {
  short *score;
  short *best;
  int *last;
  int min, max;
  int best_diag;
} DIAGONAL_SCORE;


/* PROFILE contains a psi-blast type profile */

typedef struct profile {
  char *name;             /* The name of the profile (NOT the filename !) */
  int length, width;      /* length = length of the underlying sequence. width = # of amino-acid types */
  char *seq;              /* the underlying protein sequence (containing "length" residues) */
  char *residue;          /* char[i] array which says which amino-acid is associated with column i in the profile */
  int **profile;          /* profile[j][i] gives the score for matching amino-acid i with row j in the profile */
  int max_score;          /* the max-scoring entry */
  double K, L, H, s;      /* default statistical parameters for camparisons with proteins of standard composition */       
  PREPROCESS *pre;
}
PROFILE;

/* BACKTRACK defines the types of move possible when backtracking through the path matrix */

typedef enum { STOP, HORIZ, VERT, DIAG, FRAME1, FRAME2 } BACKTRACK;


PROFILE *NextProfile( FILE *fp );
PROFILE *ReadProfile( FILE *fp );
void FreeProfile( PROFILE *pro );
PROFILE *ShuffleProfile( PROFILE *pro, int *seed );


/* DNA - profile alignment functions */

PROFILE_ALIGNMENT *DnaBacktrack( char **path, int I, int J, int *length );

void PrintProfileAlignment( FILE *fp, int count, PROFILE *pro, SEQUENCE *interlaced, 
			PROFILE_ALIGNMENT *align, int len, int score, char strand, double evalue,
			double percent, double K, double L, double H, double alpha, int show_align );

PROFILE_ALIGNMENT *AlignIseqToProfile( SEQUENCE *interlaced, PROFILE *pro, int pro_start, int pro_extend, 
				   int frameshift, int banded, int dstart, int dstop, int *score, int *length, double *percent );
int FastScoreIseqToProfile( SEQUENCE *interlaced, PROFILE *pro, char *proname, int pro_start, int pro_extend, 
			    int frameshift, int strand, PREPROCESS *pre, double ungappedFilter, int dbsize, double ethresh, double default_K, double default_L, int align, FILE *outfp );
  

void PrintProfile( FILE *fp, PROFILE *pro );
double PsiLambda( PROFILE *pro, SEQUENCE *seq );

double PercentIdentity( PROFILE_ALIGNMENT *align, int length, SEQUENCE *seq, PROFILE *pro );

/* protein - profile alignment functions */

PROFILE_ALIGNMENT *ProteinBacktrack( unsigned int **path, int I, int J, int *length );

PROFILE_ALIGNMENT *AlignProteinToProfile( SEQUENCE *protein, PROFILE *pro, int start, int extend, 
				      int *score, int *length, double *percent );

int ScoreProteinToProfile( SEQUENCE *protein, PROFILE *pro, char *proname, 
			   int pro_start, int pro_extend, 
			   int banded, int band_start, int band_stop, int align, int dbsize, double ethresh, double default_K, double default_L, FILE *outfp );
int FastScoreProteinToProfile( SEQUENCE *protein, PROFILE *pro, char *proname, int pro_start, int pro_extend, 
			       PREPROCESS *pre, int align, int dbsize, double ethresh, double default_K, double default_L, FILE *outfp );

void PrintProteinProfileAlignment( FILE *fp, int count, PROFILE *pro, SEQUENCE *protein, 
			       PROFILE_ALIGNMENT *align, int len, int score, double evalue, double percent, double K, double L, double H, double alpha, int show_align );


int  *ResidueMapping(void);
void FreePreProcess( PREPROCESS *pre );
DIAGONAL_SCORE *DiagonalScore( SEQUENCE *seq, PREPROCESS *pre );
DIAGONAL_SCORE *InterlacedDiagonalScore( SEQUENCE *seq, PREPROCESS *pre);
void FreeDiagonalScore( DIAGONAL_SCORE *ds );
int FindFastPvalues( SCORE *hit, int hits, double *K, double *L, int *positives, double *length );

PROFILE *ReadBinaryProfile( FILE *fp );
int WriteBinaryProfile( PROFILE *pro, FILE *fp );
int WritePreProcessedProfile( PREPROCESS *pre, FILE *fp );
PREPROCESS *ReadPreProcesedProfile( FILE *fp );
int DeleteWords( PREPROCESS *pre, int position_frequency );

int EvaluateProteinMatchSignificance( SEQUENCE *protein, PROFILE *pro, int A, int B, int score, int dbsize, double ethresh, double default_K, double default_L, int show_align, FILE *out_fp );
int EvaluateDNAMatchSignificance( SEQUENCE *interlaced, PROFILE *pro, int strand, int A, int B, int F, int banded, int dstart, int dstop, int score, int dbsize, double ethresh, double default_K, double default_L, int show_align, FILE *out_fp );

int KarlinAltschulProfileStatistics( PROFILE *pro, double *freq, double *lambda, double *Kminus, double *Kplus, double *H, double *r, double *s );
double *GetPsiH( PROFILE *pro, double *freq, int *hmin, int *hmax, double *mean );
int SetDefaultParameters( PROFILE *pro );

PROFILE *SeqToProfile( SEQUENCE *seq, int **matrix );
PROFILE *CheapSeqToProfile( SEQUENCE *seq, int **matrix );
void FreeCheapProfile( PROFILE *cheap );
PREPROCESS  *PreProcessProfile( PROFILE *pro, int threshold, int diag_threshold, int max_dist, int window, int position_frequency );

void IterativelyIncreaseGapPenalty( SEQUENCE *seq, PROFILE *pro, int *A, int *B, double lambda0, double K0, double H, double s, double *alpha, int *sw_score );

#endif