GtScoreMatrix* gt_score_matrix_new(GtAlphabet *alphabet)
{
GtScoreMatrix *sm;
gt_assert(alphabet);
sm = gt_malloc(sizeof (GtScoreMatrix));
sm->alphabet = gt_alphabet_ref(alphabet);
sm->dimension = gt_alphabet_size(alphabet);
gt_array2dim_calloc(sm->scores, sm->dimension, sm->dimension);
return sm;
}
static GtBaseQualDistr* hcr_base_qual_distr_new_from_file(FILE *fp,
GtAlphabet *alpha)
{
GtBaseQualDistr *bqd;
char read_char_code;
GtUchar cur_char_code;
unsigned char cur_qual;
unsigned alpha_size,
min_qual = HCR_HIGHESTQUALVALUE,
max_qual = HCR_LOWESTQUALVALUE;
GtUword numofleaves,
i;
GtUint64 cur_freq;
GT_UNUSED size_t read,
one = (size_t) 1;
alpha_size = gt_alphabet_size(alpha);
bqd = gt_malloc(sizeof (GtBaseQualDistr));
gt_array2dim_calloc(bqd->distr, HCR_HIGHESTQUALVALUE + 1UL, alpha_size)
bqd->ncols = alpha_size;
bqd->nrows = HCR_HIGHESTQUALVALUE + 1U;
bqd->qual_offset = HCR_LOWESTQUALVALUE;
bqd->wildcard_indx = alpha_size - 1;
read = gt_xfread_one(&numofleaves, fp);
gt_assert(read == one);
for (i = 0; i < numofleaves; i++) {
read = gt_xfread_one(&read_char_code, fp);
gt_assert(read == one);
read = gt_xfread_one(&cur_qual, fp);
gt_assert(read == one);
read = gt_xfread_one(&cur_freq, fp);
gt_assert(read == one);
cur_char_code = gt_alphabet_encode(alpha, read_char_code);
if (cur_char_code == (GtUchar) WILDCARD)
gt_safe_assign(cur_char_code, bqd->wildcard_indx);
bqd->distr[cur_qual][cur_char_code] = cur_freq;
if ((unsigned) cur_qual > max_qual)
max_qual = cur_qual;
if ((unsigned) cur_qual < min_qual)
min_qual = cur_qual;
}
bqd->min_qual = min_qual;
bqd->max_qual = max_qual;
hcr_base_qual_distr_trim(bqd);
return bqd;
}
static GtBaseQualDistr* hcr_base_qual_distr_new(GtAlphabet *alpha,
GtQualRange qrange)
{
GtBaseQualDistr *bqd;
bqd = gt_calloc((size_t) 1, sizeof (GtBaseQualDistr));
gt_array2dim_calloc(bqd->distr, HCR_HIGHESTQUALVALUE + 1UL,
gt_alphabet_size(alpha));
bqd->ncols = gt_alphabet_size(alpha);
bqd->nrows = HCR_HIGHESTQUALVALUE + 1U;
bqd->qual_offset = HCR_LOWESTQUALVALUE;
bqd->wildcard_indx = gt_alphabet_size(alpha) - 1;
bqd->min_qual = HCR_HIGHESTQUALVALUE;
bqd->max_qual = HCR_LOWESTQUALVALUE;
gt_safe_assign(bqd->qrange_start, qrange.start);
gt_safe_assign(bqd->qrange_end, qrange.end);
bqd->alpha = alpha;
return bqd;
}
static void hcr_base_qual_distr_trim(GtBaseQualDistr *bqd)
{
if (bqd->min_qual != 0) {
GtUint64 **distr_trimmed;
unsigned nrows_new,
i,
j;
nrows_new = bqd->max_qual - bqd->min_qual + 1;
gt_array2dim_calloc(distr_trimmed, (GtUword) nrows_new, bqd->ncols);
for (i = 0; i < nrows_new; i++)
for (j = 0; j < bqd->ncols; j++)
distr_trimmed[i][j] = bqd->distr[i + bqd->min_qual][j];
gt_array2dim_delete(bqd->distr);
bqd->distr = distr_trimmed;
bqd->nrows = nrows_new;
bqd->qual_offset = bqd->min_qual;
}
}
static int gt_genomediff_runner(int argc, const char **argv,
int parsed_args, void *tool_arguments,
GtError *err)
{
bool mirrored = false;
int had_err = 0,
i;
GtEncseq *encseq = NULL;
GtGenomediffArguments *arguments = tool_arguments;
GtLogger *logger;
GtShuUnitFileInfo *unit_info = NULL;
GtTimer *timer = NULL;
gt_error_check(err);
gt_assert(arguments);
logger = gt_logger_new(arguments->verbose,
GT_LOGGER_DEFLT_PREFIX,
stdout);
gt_assert(logger);
for (i = parsed_args; i < argc; i++) {
gt_str_array_add_cstr(arguments->filenames, argv[i]);
}
if (gt_showtime_enabled()) {
timer = gt_timer_new_with_progress_description("start");
gt_timer_start(timer);
gt_assert(timer);
}
if (arguments->with_units) {
gt_logger_log(logger, "unitfile option set, filename is %s\n",
gt_str_get(arguments->unitfile));
}
if (timer != NULL)
gt_timer_show_progress(timer, "start shu search", stdout);
if (gt_str_array_size(arguments->filenames) > 1UL) {
GtEncseqEncoder *ee = gt_encseq_encoder_new();
gt_encseq_encoder_set_timer(ee, timer);
gt_encseq_encoder_set_logger(ee, logger);
/* kr only makes sense for dna, so we can check this already with ee */
gt_encseq_encoder_set_input_dna(ee);
had_err = gt_encseq_encoder_encode(ee, arguments->filenames,
gt_str_get(arguments->indexname), err);
gt_encseq_encoder_delete(ee);
}
else {
gt_str_append_str(arguments->indexname,
gt_str_array_get_str(arguments->filenames, 0));
if (arguments->with_esa || arguments->with_pck) {
GtStr *current_line = gt_str_new();
FILE *prj_fp;
const char *buffer;
char **elements = NULL;
prj_fp = gt_fa_fopen_with_suffix(gt_str_get(arguments->indexname),
GT_PROJECTFILESUFFIX,"rb",err);
if (prj_fp == NULL)
had_err = -1;
while (!had_err && gt_str_read_next_line(current_line, prj_fp) != EOF) {
buffer = gt_str_get(current_line);
if (elements != NULL) {
gt_free(elements[0]);
gt_free(elements[1]);
}
gt_free(elements);
elements = gt_cstr_split(buffer, '=');
gt_log_log("%s", elements[0]);
if (strcmp("mirrored", elements[0]) == 0) {
gt_log_log("%s", elements[1]);
if (strcmp("1", elements[1]) == 0) {
mirrored = true;
gt_log_log("sequences are treated as mirrored");
}
}
gt_str_reset(current_line);
}
gt_str_delete(current_line);
if (elements != NULL) {
gt_free(elements[0]);
gt_free(elements[1]);
}
gt_free(elements);
gt_fa_xfclose(prj_fp);
}
}
if (!had_err) {
GtEncseqLoader *el = gt_encseq_loader_new_from_options(arguments->loadopts,
err);
if (mirrored)
gt_encseq_loader_mirror(el);
encseq =
gt_encseq_loader_load(el, gt_str_get(arguments->indexname), err);
gt_encseq_loader_delete(el);
}
if (encseq == NULL)
had_err = -1;
if (!had_err) {
unit_info = gt_shu_unit_info_new(encseq);
if (arguments->with_units)
had_err = gt_shu_unit_file_info_read(arguments->unitfile, unit_info,
logger, err);
}
if (!had_err) {
uint64_t **shusums = NULL;
if (arguments->with_esa || arguments->with_pck) {
shusums = gt_genomediff_shulen_sum(arguments, unit_info,
logger, timer, err);
if (shusums == NULL)
had_err = -1;
}
else {
const bool doesa = true;
GenomediffInfo gd_info;
Suffixeratoroptions sopts;
sopts.beverbose = arguments->verbose;
sopts.indexname = arguments->indexname;
sopts.db = NULL;
sopts.encopts = NULL;
sopts.genomediff = true;
sopts.inputindex = arguments->indexname;
sopts.loadopts = arguments->loadopts;
sopts.showprogress = false;
sopts.idxopts = arguments->idxopts;
gt_assert(unit_info != NULL);
gt_array2dim_calloc(shusums, unit_info->num_of_genomes,
unit_info->num_of_genomes);
gd_info.shulensums = shusums;
gd_info.unit_info = unit_info;
had_err = gt_runsuffixerator(doesa, &sopts, &gd_info, logger, err);
}
if (!had_err && shusums != NULL) {
had_err = gt_genomediff_kr_calc(shusums, arguments, unit_info,
arguments->with_pck, logger, timer, err);
gt_array2dim_delete(shusums);
}
}
if (timer != NULL) {
gt_timer_show_progress_final(timer, stdout);
gt_timer_delete(timer);
}
gt_logger_delete(logger);
gt_encseq_delete(encseq);
gt_shu_unit_info_delete(unit_info);
return had_err;
}
static int gt_gdiffcalc_runner(int argc, const char **argv, int parsed_args,
void *tool_arguments, GT_UNUSED GtError *err)
{
GtGenomediffArguments *arguments = tool_arguments;
int had_err = 0, i;
GtUword lcounter = 0, zcounter = 0;
double **shusums = NULL;
GtEncseq *encseq = NULL;
GtLogger *logger;
GtShuUnitFileInfo *unit_info = NULL;
GtTimer *timer = NULL;
gt_error_check(err);
gt_assert(arguments);
logger = gt_logger_new(arguments->verbose,
GT_LOGGER_DEFLT_PREFIX,
stdout);
gt_assert(logger);
for (i = parsed_args; i < argc; i++) {
gt_str_array_add_cstr(arguments->filenames, argv[i]);
}
if (gt_showtime_enabled()) {
timer = gt_timer_new_with_progress_description("load encseq");
gt_timer_start(timer);
gt_assert(timer);
}
if (arguments->with_units) {
gt_logger_log(logger, "unitfile option set, filename is %s\n",
gt_str_get(arguments->unitfile));
}
if (!had_err) {
GtEncseqLoader *el = gt_encseq_loader_new_from_options(arguments->loadopts,
err);
encseq =
gt_encseq_loader_load(el, gt_str_get(arguments->indexname), err);
gt_encseq_loader_delete(el);
}
if (encseq == NULL)
had_err = -1;
if (timer != NULL)
gt_timer_show_progress(timer, "load units", stdout);
if (!had_err) {
unit_info = gt_shu_unit_info_new(encseq);
if (arguments->with_units)
had_err = gt_shu_unit_file_info_read(arguments->unitfile, unit_info,
logger, err);
}
if (timer != NULL)
gt_timer_show_progress(timer, "read table", stdout);
if (!had_err) {
GtIO *table_file = NULL;
GtTokenizer *tokenizer = NULL;
GtStr *line = NULL;
gt_assert(unit_info != NULL);
gt_array2dim_calloc(shusums, unit_info->num_of_genomes,
unit_info->num_of_genomes);
table_file = gt_io_new(gt_str_array_get(arguments->filenames, 0), "r");
tokenizer = gt_tokenizer_new(table_file);
line = gt_tokenizer_get_token(tokenizer);
while (line != NULL && !had_err) {
char *cline = gt_str_get(line);
char *elem = strtok(cline, ";");
zcounter = 0;
while (elem != NULL && !had_err) {
if (*elem != '#') {
if (1 != sscanf(elem, "%lf",
&shusums[lcounter][zcounter])) {
had_err = 1;
gt_error_set(err, "couldn't scan");
break;
}
gt_logger_log(logger,"wert: %lf", shusums[lcounter][zcounter]);
zcounter++;
}
else {
gt_logger_log(logger, "name: %s", elem++);
}
elem = strtok(NULL, ";");
}
gt_tokenizer_next_token(tokenizer);
gt_str_delete(line);
line = gt_tokenizer_get_token(tokenizer);
lcounter++;
gt_logger_log(logger, "line "GT_WD"", lcounter);
}
}
if (!had_err) {
GtUword num_of_seq, file_idx, seq_idx, startpos;
GT_UNUSED GtUword oldpos = 0;
gt_assert(unit_info != NULL);
gt_assert(lcounter == zcounter);
gt_assert(lcounter == unit_info->num_of_genomes);
num_of_seq = gt_encseq_num_of_sequences(unit_info->encseq);
for (seq_idx = 0; seq_idx < num_of_seq; seq_idx++) {
startpos = gt_encseq_seqstartpos(unit_info->encseq, seq_idx);
file_idx = gt_encseq_filenum(unit_info->encseq, startpos);
gt_log_log("seq: "GT_WU" starts at: "GT_WU"\n"
"belonges to file: "GT_WU" which is part of genome: %s",
seq_idx, startpos, file_idx,
gt_str_array_get(unit_info->genome_names,
unit_info->map_files[file_idx]));
gt_assert(oldpos <= startpos);
oldpos = startpos;
}
}
if (!had_err && shusums != NULL) {
had_err = gt_genomediff_calculate_div_from_avg(shusums, arguments,
unit_info,
logger, timer, err);
gt_array2dim_delete(shusums);
}
if (timer != NULL) {
gt_timer_show_progress_final(timer, stdout);
gt_timer_delete(timer);
}
gt_logger_delete(logger);
gt_encseq_delete(encseq);
gt_shu_unit_info_delete(unit_info);
return had_err;
}
int mg_computepath(CombinedScoreMatrixEntry **combinedscore_matrix,
HitInformation *hit_information,
unsigned long rows,
unsigned long contig_len,
ParseStruct *parsestruct_ptr, GtError * err)
{
int had_err = 0;
/* Initialisieren der Matrix fuer die Pfadberechnung */
PathMatrixEntry **path_matrix;
/* i: Zaehlvariable fuer die Matrix-Zeilen; k: Zaehlvariable Precursors
(von 0 bis max 2) maxpath_frame: Speichern des vorherigen Frames von
dem der max-Wert berechnet wird */
unsigned short row_index = 0,
precursor_index = 0,
precursors_row = 0,
maxpath_frame = 0;
/* Position in der Query-DNA */
unsigned long column_index = 0;
/* Variablen fuer den aktuellen Frame, den vorherigen Frame(speichert
einen Wert aus precursors[], die Zeile des vorherigen Frames, GtArray
mit den Precursors-Frames */
short current_frame = 0,
precursors_frame = 0,
precursors[NUM_PRECURSORS];
/* q ist der Wert, der bei Aus- oder Eintreten in ein Gen auf dem
Forward- bzw. Reverse-Strang berechnet wird */
double q = ARGUMENTSSTRUCT(leavegene_value),
max_new = 1,
max_old = 1;
/* Speicherreservierung fuer die Path-Matrix - Groesse entsprechend der
CombinedScore-Matrix */
gt_array2dim_calloc(path_matrix, 7, contig_len);
gt_error_check(err);
/* fuer die erste Spalte der Path-Matrix wird die erste Spalte der
CombinedScore-Matrix uebernommen */
for (row_index = 0; row_index < rows; row_index++)
{
path_matrix[row_index][0].score =
combinedscore_matrix[row_index][0].matrix_score;
path_matrix[row_index][0].path_frame = row_index;
}
/* Spaltenweise Berechnung des opt. Pfades */
for (column_index = 1; column_index < contig_len; column_index++)
{
for (row_index = 0; row_index < rows; row_index++)
{
/* Zaehlvariable fuer die Zeile wird umgerechnet in den entsprechenden
Leserahmen */
current_frame = get_current_frame(row_index);
/* Aufruf der Methode zum Berechnen der moeglichen Leserahmen anhand von
aktuellem Leserahmen und der Query-DNA-Sequenz */
compute_precursors(current_frame,
column_index,
precursors);
/* der max-Wert der moeglichen Vorgaenger wird berechnet */
for (precursor_index = 0;
precursor_index < NUM_PRECURSORS
&& (precursors[precursor_index] != UNDEFINED);
++precursor_index)
{
/* aktueller Vorgaengerleserahmen - es gibt max. 3 moegliche
Vorgaenger */
precursors_frame = precursors[precursor_index];
/* Vorgaengerleserahmen wird umgerechnet in die entsprechende
Matrix-Zeile */
precursors_row = get_matrix_row(precursors_frame);
/* der DP-Algo umfasst 3 moegliche Faelle
1. Fall: Wechsel vom Reversen- auf den Forward-Strang bzw.
umgekehrt */
if ((current_frame < 0 && precursors_frame > 0) ||
(current_frame > 0 && precursors_frame < 0))
{
max_new = path_matrix[precursors_row][column_index-1].score +
combinedscore_matrix[row_index][column_index].matrix_score
+ 2*q;
}
/* 2. Fall: Einfacher Wechsel des Leserahmens, also von + zu +
bzw.- zu - */
else if (current_frame != 0 && precursors_frame != current_frame)
{
max_new = path_matrix[precursors_row][column_index-1].score +
combinedscore_matrix[row_index][column_index].matrix_score
+ q;
}
/* 3. Fall: Leserahmen wird beibehalten bzw. Wechsel von kodierend zu
nicht-kodierend oder umgekehrt */
else
{
max_new = path_matrix[precursors_row][column_index-1].score +
combinedscore_matrix[row_index][column_index]
.matrix_score;
}
/* Bestimmen des Max-Wertes der max. 3 Moeglichkeiten und Speichern der
Zeile, von der der Max-Wert stammt */
if (gt_double_compare(max_new, max_old) > 0)
{
max_old = max_new;
maxpath_frame = precursors_row;
}
}
/* Speichern des Max-Wertes und der "Vorgaenger"-Zeile;
zuruecksetzen der Variablen */
path_matrix[row_index][column_index].score = max_old;
path_matrix[row_index][column_index].path_frame = maxpath_frame;
max_new = DBL_MIN;
max_old = DBL_MIN;
maxpath_frame = 0;
}
}
/* Aufruf der Methode zur Genvorhersage */
had_err = mg_compute_gene_prediction(combinedscore_matrix,
path_matrix,
contig_len,
hit_information,
parsestruct_ptr, err);
gt_array2dim_delete(path_matrix);
return had_err;
}
