7cc2ad1ed09ac992db9cd5d783e871cbdba4dba3 larrym Mon Jul 19 00:25:27 2010 -0700 add dnaMotifBitScoreWithMark0Bg diff --git src/inc/dnaMotif.h src/inc/dnaMotif.h index 54aecc4..6d4f7b1 100644 --- src/inc/dnaMotif.h +++ src/inc/dnaMotif.h @@ -1,123 +1,126 @@ /* dnaMotif.h was originally generated by the autoSql program, which also * generated dnaMotif.c and dnaMotif.sql. This header links the database and * the RAM representation of objects. */ #ifndef DNAMOTIF_H #define DNAMOTIF_H #ifndef DNAUTIL_H #include "dnautil.h" #endif struct dnaMotif /* A gapless DNA motif */ { struct dnaMotif *next; /* Next in singly linked list. */ char *name; /* Motif name. */ int columnCount; /* Count of columns in motif. */ float *aProb; /* Probability of A's in each column. */ float *cProb; /* Probability of C's in each column. */ float *gProb; /* Probability of G's in each column. */ float *tProb; /* Probability of T's in each column. */ }; struct dnaMotif *dnaMotifCommaIn(char **pS, struct dnaMotif *ret); /* Create a dnaMotif out of a comma separated string. * This will fill in ret if non-null, otherwise will * return a new dnaMotif */ void dnaMotifFree(struct dnaMotif **pEl); /* Free a single dynamically allocated dnaMotif such as created * with dnaMotifLoad(). */ void dnaMotifFreeList(struct dnaMotif **pList); /* Free a list of dynamically allocated dnaMotif's */ void dnaMotifOutput(struct dnaMotif *el, FILE *f, char sep, char lastSep); /* Print out dnaMotif. Separate fields with sep. Follow last field with lastSep. */ #define dnaMotifTabOut(el,f) dnaMotifOutput(el,f,'\t','\n'); /* Print out dnaMotif as a line in a tab-separated file. */ #define dnaMotifCommaOut(el,f) dnaMotifOutput(el,f,',',','); /* Print out dnaMotif as a comma separated list including final comma. */ float dnaMotifSequenceProb(struct dnaMotif *motif, DNA *dna); /* Return probability of dna according to motif. Make sure * motif is probabalistic (with call to dnaMotifMakeProbabalistic * if you're not sure) before calling this. */ char dnaMotifBestStrand(struct dnaMotif *motif, DNA *dna); /* Figure out which strand of DNA is better for probabalistic motif. */ double dnaMotifBitScore(struct dnaMotif *motif, DNA *dna); /* Return logBase2-odds score of dna given a probabalistic motif. */ +double dnaMotifBitScoreWithMark0Bg(struct dnaMotif *motif, DNA *dna, double mark0[5]); +/* Return logBase2-odds score of dna given a probabalistic motif and using a 0-order markov model for the background. */ + double dnaMotifBitScoreWithMarkovBg(struct dnaMotif *motif, DNA *dna, double mark2[5][5][5]); /* Return logBase2-odds score of dna given a probabalistic motif using a 2nd-order markov model for the background. motif and markd2 must be in log2 format. Seq must contain an extra two bases at the front (i.e. we start scoring from dna + 2). */ void dnaMotifMakeLog2(struct dnaMotif *motif); void dnaMotifNormalize(struct dnaMotif *motif); /* Make all columns of motif sum to one. */ boolean dnaMotifIsScoreBased(struct dnaMotif *motif); /* Return TRUE if dnaMotif is score-based (which we decide by * the presense of negative values. */ void dnaMotifScoreToProb(struct dnaMotif *motif); /* Convert motif that is log-odds score based to motif * that is probability based. This assumes that the * background distribution is simple: 25% for each base */ void dnaMotifMakeProbabalistic(struct dnaMotif *motif); /* Change motif, which may be score or count based, to * probabalistic one, where each column adds to 1.0 */ double dnaMotifBitsOfInfo(struct dnaMotif *motif, int pos); /* Return bits of information at position. */ void dnaMotifPrintProb(struct dnaMotif *motif, FILE *f); /* Print DNA motif probabilities. */ void dnaMotifToLogoPs2(struct dnaMotif *motif, double widthPerBase, double height, double minHeight, char *fileName); /* Write logo corresponding to motif to postScript file, with extended options. minHeight * is the minimum height that is excluded from information content scaling. This allows * something to show up in columns with very little information content. Setting this * to be the same as height creates an frequency-based logo. */ void dnaMotifToLogoPs(struct dnaMotif *motif, double widthPerBase, double height, char *fileName); /* Write logo corresponding to motif to postScript file. */ void dnaMotifToLogoPsW(struct dnaMotif *motif, double widthPerBase, double width, double height, char *fileName); /* Write logo corresponding to motif to postScript file. * use the whole window width, rather than just incrementing * by widthPerBase */ void dnaMotifToLogoPng( struct dnaMotif *motif, /* Motif to draw. */ double widthPerBase, /* Width of each base. */ double height, /* Max height. */ char *gsExe, /* ghostscript executable, NULL for default */ char *tempDir, /* temp dir , NULL for default */ char *fileName); /* output png file name. */ /* Write logo corresponding to motif to png file. */ void dnaMotifToLogoPGM( struct dnaMotif *motif, /* Motif to draw. */ double widthPerBase, /* Width of each base. */ double width, /* Max width. */ double height, /* Max height. */ char *gsExe, /* ghostscript executable, NULL for default */ char *tempDir, /* temp dir , NULL for default */ char *fileName); /* output png file name. */ /* Write logo corresponding to motif to PGM file. */ #endif /* DNAMOTIF_H */