Score logl_positive_selection(char * ref,char * diff,RandomModel * rm,CodonTable *ct,DnaProbMatrix * dm)
{
int i;
Score s = 0;
char ref_aa;
char diff_aa;
/* we have to assess this position having changed */
for(i=0;i<3;i++) {
s += Probability2Score(dm->prob[base_from_char(ref[i])][base_from_char(diff[i])]);
}
/* if the position has not changed, then we know it would have been selected */
ref_aa = aminoacid_from_seq(ct,ref);
diff_aa = aminoacid_from_seq(ct,diff);
if( ref_aa == diff_aa ) {
return s;
}
/* else we add the probability of seeing this amino acid*/
s += Probability2Score(rm->aminoacid[diff_aa-'A']);
return s;
}
Score logl_pseudogene(char * ref,char * diff,DnaProbMatrix * dm)
{
int i;
Score s = 0;
for(i=0;i<3;i++) {
s += Probability2Score(dm->prob[base_from_char(ref[i])][base_from_char(diff[i])]);
}
return s;
}
codon codon_from_seq(char * seq)
{
base one;
base two;
base three;
one = base_from_char(*seq);
two = base_from_char(*(seq+1));
three = base_from_char(*(seq+2));
return one*25+two*5+three;
}
DnaNumber dna_number_from_string(char * str,int nmer_size)
{
int i;
int base = 1;
DnaNumber out;
int forward;
int backward;
out.flipped = 2;
out.number = 0;
for(i=0;i<nmer_size-1;i++)
base *= 4;
for(i=0;i<nmer_size;i++) {
forward = base_from_char(str[i]);
backward = complement_base(base_from_char(str[nmer_size-1-i]));
if( forward == BASE_N || backward == BASE_N ) {
return out;
}
if( forward > backward ) {
out.flipped = 0;
break;
}
if( backward > forward ) {
out.flipped = 1;
break;
}
}
assert(out.flipped != 2);
if( out.flipped == 0 ) {
for(i=0;i<nmer_size;i++) {
out.number += base * base_from_char(str[i]);
base = base / 4;
}
} else {
for(i=0;i<nmer_size;i++) {
/* fprintf(stderr,"For position %d, [%d], using %d [%c]as complemented base\n",i,base,complement_base(base_from_char(str[nmer_size-1-i])),str[nmer_size-1-i]); */
out.number += base * complement_base(base_from_char(str[nmer_size-1-i]));
base = base / 4;
}
}
return out;
}
int base4_codon_from_seq(char * seq)
{
base one;
base two;
base three;
one = base_from_char(*seq);
two = base_from_char(*(seq+1));
three = base_from_char(*(seq+2));
if( one == BASE_N || two == BASE_N || three == BASE_N)
return 64;
else return one*16+two*4+three;
}
static void read_int22(int use_dna_params, tab8_t *t)
{
fill8(t, NOT_A_NUMBER);
FILE *f = parfile("int22", use_dna_params);
look_for_line_containing(f, "5' ------> 3'");
char buf[MAXLINE+1];
while (fgets(buf, MAXLINE, f)) {
if (strlen(buf) >= MAXLINE)
die("read_int22: line too long");
if (is_only_whitespace(buf))
continue;
look_for_line_containing(f, "5' ------> 3'");
char a[4];
if (!(fgets(buf, MAXLINE, f) && sscanf(buf, " %c \\/ \\_/ %c", &a[0], &a[1]) == 2))
die("read_int22: couldn't read first line");
if (!(fgets(buf, MAXLINE, f) && sscanf(buf, " %c /\\ | %c", &a[2], &a[3]) == 2))
die("read_int22: couldn't read second line");
expect_line_containing(f, "3' <------ 5'");
int i, j;
base_t b[4];
for (i = 0; i < 4; i++)
b[i] = base_from_char(a[i]);
for (i = 0; i < 4; i++)
for (j = 0; j < 4; j++) {
int_t val[16];
read_next_values(f, val, 16);
int k, l;
for (k = 0; k < 4; k++)
for (l = 0; l < 4; l++)
(*t)[b[0]][b[1]][b[2]][b[3]][i][k][j][l] = val[4*k+l];
}
}
fclose(f);
}
boolean is_non_ambiguous_codon_seq(char * seq)
{
if( *seq == '\0' || *(seq+1) == '\0' || *(seq+2) == '\0') {
warn("Attempting to find a codon number is something less than 3 bases long!");
return FALSE;
}
if( base_from_char(*(seq++)) == BASE_N)
return FALSE;
if( base_from_char(*(seq++)) == BASE_N)
return FALSE;
if( base_from_char(*(seq)) == BASE_N)
return FALSE;
return TRUE;
}
static void read_small_loop(char *buf, base_t seq[], int_t *val, int n)
{
char *b = strtok(buf, whitespace);
int i;
for (i = 0; i < n; i++)
seq[i] = base_from_char(b[i]);
b = strtok(0, whitespace);
*val = value_from_string(b);
}
static void read_six_bases(const char *buf, base_t b[6])
{
char a[6];
if (sscanf(buf, " Y%c Y%c Y%c Y%c Y%c Y%c",
&a[0], &a[1], &a[2], &a[3], &a[4], &a[5]) != 6)
die("read_six_bases: error");
int i;
for (i = 0; i < 6; i++)
b[i] = base_from_char(a[i]);
}
static void read_four_bases(const char *buf, base_t b[4])
{
char a[4];
const char *fmt = strchr(buf,'X') ? " %cX %cX %cX %cX" : " %c %c %c %c";
if (sscanf(buf, fmt, &a[0], &a[1], &a[2], &a[3]) != 4)
die("read_four_bases: error");
int i;
for (i = 0; i < 4; i++)
b[i] = base_from_char(a[i]);
}
static void read_twelve_bases(const char *buf, base_t b[12])
{
char a[12];
if (sscanf(buf, " %c %c %c %c %c %c %c %c %c %c %c %c",
&a[0], &a[1], &a[2], &a[3],
&a[4], &a[5], &a[6], &a[7],
&a[8], &a[9], &a[10], &a[11]) != 12)
die("read_twelve_bases: error");
int i;
for (i = 0; i < 12; i++)
b[i] = base_from_char(a[i]);
}
double nocds_from_ambiguous_codon(char * codon,double * codon_freq_array)
{
int factor = 1;
int one;
int two;
int three;
int i,j,k;
double ret = 0.0;
one = base_from_char(*codon == '-' ? 'N' : *codon);
two = base_from_char(*(codon+1) == '-' ? 'N' : *(codon+1));
three = base_from_char(*(codon+2) == '-' ? 'N' : *(codon+2));
if(one == BASE_N)
factor *= 4;
if(two == BASE_N)
factor *= 4;
if(three == BASE_N)
factor *= 4;
for(i=0;i<4;i++)
for(j=0;j<4;j++)
for(k=0;k<4;k++)
if( (one == i || one == BASE_N) && (two == j || two == BASE_N) && (three == k || three == BASE_N)) {
ret += codon_freq_array[i*16+j*4+k];
}
ret = ret / factor;
if( ret < 0.0000000000000001 ) {
warn("For codon %c%c%c we have a frequency of %g",*codon,*(codon+1),*(codon+2),ret);
ret = 0.0000000000000001;
}
return ret;
}
Score logl_negative_selection(char * ref,char * diff,ThreeStateUnit * unit,CodonTable * ct,DnaProbMatrix * dm)
{
int i;
Score s = 0;
char ref_aa;
char diff_aa;
/* we have to assess this position having changed */
for(i=0;i<3;i++) {
s += Probability2Score(dm->prob[base_from_char(ref[i])][base_from_char(diff[i])]);
}
/* if the position has not changed, then we know it could not have been selected */
ref_aa = aminoacid_from_seq(ct,ref);
diff_aa = aminoacid_from_seq(ct,diff);
if( ref_aa == diff_aa ) {
return s;
}
/* else we add the difference in probability between the two amino acids */
/*
fprintf(stdout,"%c vs %c has %d plays %d for total of %d\n",ref_aa,diff_aa,
Probability2Score(unit->match_emission[ref_aa-'A']),
Probability2Score(unit->match_emission[diff_aa-'A']),
Probability2Score(unit->match_emission[diff_aa-'A']) - Probability2Score(unit->match_emission[ref_aa-'A'])
);
*/
s += Probability2Score(unit->match_emission[diff_aa-'A']) - Probability2Score(unit->match_emission[ref_aa-'A']);
return s;
}
boolean read_base_GeneConsensus(double * base_array,char* line,FILE * ifp)
{
boolean ret = TRUE;
int b;
char * base;
char * number;
if( strwhitestartcmp(line,"begin",spacestr) != 0 || strstr(line,"consensus") == NULL ) {
warn("In reading base GeneConsensus line, got no 'begin consensus' tag [%s]",line);
return FALSE;
}
while( fgets(line,MAXLINE,ifp) != NULL ) {
if( line[0] == '#' )
continue;
if( strwhitestartcmp(line,"end",spacestr) == 0 )
break;
base = strtok(line,spacestr);
number = strtok(NULL,spacestr);
if( base == NULL ) {
warn("Found an uncommented line in base consensus with no leading base word");
continue;
}
if( number == NULL ) {
warn("For base %s, no number found",base);
ret = FALSE;
continue;
}
if( strlen(base) > 1 || (b=base_from_char(*base)) == BASE_N ) {
warn("Could not interpret %s as an actual DNA base in read_base_GeneConsensus");
ret = FALSE;
continue;
}
base_array[b]= atof(number);
}
return ret;
}
Score SpliceSiteModel_score(SpliceSiteModel * ssm,char * seq)
{
int len;
int i;
int score;
char * be = seq;
base b;
/* check I have enough sequence */
/* fprintf(stderr,"Being passed sequence %c%c%c\n",seq[0],seq[1],seq[2]); */
/* first calculate the CC score */
score = score_from_ComplexConsensi(seq- ssm->offset - ssm->pre_splice_site,ssm->cc);
/* now move over the random score */
/* random score is subtracted - ie divided */
/* out from the model */
len = ssm->start_random - ssm->stop_random +1;
for(i=0,seq = seq - ssm->start_random+1;i<len;i++,seq++) {
if( *seq == '\0' ) {
warn("You are attempting to score an impossible base (%d from SS) [%s] in a splice site",(int)(seq - be),be);
return NEGI;
}
b = base_from_char(*seq);
score -= ssm->rmds->base[b];
}
/* this is for the possibility of errors/non splice consensus etc */
if( score < ssm->error_pos )
score = ssm->error_pos;
return score;
}
DnaHmmProbUnit * new_DnaHmmProbUnit_from_ColumnCount_ungapped(ColumnCount * cc,double simple_pseudocount)
{
DnaHmmProbUnit * out;
double total;
char base[] = "ATGC";
int i;
out = DnaHmmProbUnit_alloc();
for(i=0,total = 0.0;i<4;i++)
total += (cc->count[base[i]-'A'] + simple_pseudocount);
for(i=0;i<4;i++)
out->match[base_from_char(base[i])] = (cc->count[base[i]-'A'] + simple_pseudocount) / total;
out->match[4] = 1.0;
for(i=0;i<DHMM_TRANSITION_LEN;i++) {
out->transition[i] = 0.0;
}
out->transition[DHMM_MATCH2MATCH] = 1.0;
return out;
}
int base_number_func(int type,void * data,char * seq)
{
return base_from_char(*seq);
}
char char_complement_base(char c)
{
return char_from_base(complement_base(base_from_char(c)));
}