cell *
prepare_dictionary(int *argcp, char *(*argvp[]))
{
u_char *origin;
u_char *here;
u_char *xlimit;
int dict_size;
u_char *extension;
char *dictionary_file = "";
// Allocate space for the Forth dictionary and read its initial contents
origin = aln_alloc(MAXDICT, variables);
xlimit = &origin[MAXDICT];
if(*argcp < 2
|| (extension = strrchr((*argvp)[1],'.')) == NULL
|| strcmp(extension, ".dic") != 0 ) {
dictionary_file = is_readable("app.dic") ? "app.dic" : DEFAULT_EXE;
} else {
dictionary_file = (*argvp)[1];
*argcp -= 1;
*argvp += 1;
}
dict_size = read_dictionary(dictionary_file, origin, variables);
here = &origin[dict_size];
init_compiler(origin, xlimit, 0xfffe, here, xlimit, variables);
return variables;
}
int main(int argc, char **argv)
{
extern void init_dictionary(cell *up);
u_char *origin;
if (argc != 3) {
fprintf(stderr, "usage: meta input-file-name output-file-name\n");
exit(1);
}
infile = argv[1];
outfile = argv[2];
origin = aln_alloc(MAXDICT, variables);
*(token_t *)origin = 0;
init_compiler(origin, origin+MAXDICT, 0xfffe, origin, origin+MAXDICT, variables);
init_dictionary(variables);
return 0;
}
struct alignment* detect_and_read_sequences(struct alignment* aln,struct parameters* param)
{
int feature = 0;
char **input = 0;
unsigned short int* input_type = 0;
unsigned short int* input_numseq = 0;
int num_input = 0;
int i = 0;
int j = 0;
int c = 0;
int a,b;
int free_read = 1;
unsigned int numseq = get_kalign_context()->numseq;
while(free_read == 1 || param->infile[i]){
num_input++;
i++;
free_read = 0;
}
numseq = 0;
input = malloc(sizeof(char*) * num_input);
input_type = malloc(sizeof(unsigned short int) * num_input);
input_numseq = malloc(sizeof(unsigned short int) * num_input);
for (i = 0; i < num_input;i++){
input[i] = 0;
input_type[i] = 0;
input_numseq[i] = 0;
}
free_read = 0;
if(param->quiet){
c = 1;
}else{
c = 0;
}
for (i = c; i < num_input;i++){
if(!param->infile[i]){
k_printf("reading from STDIN: ");
}else{
k_printf("reading from %s: ",param->infile[i]);
}
input[i] = get_input_into_string(input[i],param->infile[i]);
if(input[i]){
free_read++;
if (byg_start("<macsim>",input[i]) != -1){
input_numseq[i] = count_sequences_macsim(input[i]);
feature = 1;
input_type[i] = 1;
}else if (byg_start("<uniprot",input[i]) != -1){
input_numseq[i] = count_sequences_uniprot(input[i]);
input_type[i] = 2;
}else if(byg_start("This SWISS-PROT",input[i]) != -1){
input_numseq[i] = count_sequences_swissprot(input[i]);
input_type[i] = 3;
}else if (byg_start("This Swiss-Prot",input[i]) != -1){
input_numseq[i] = count_sequences_swissprot(input[i]);
input_type[i] = 3;
}else if (byg_start("CLUSTAL W",input[i]) != -1){
input_numseq[i] = count_sequences_clustalw(input[i]);
input_type[i] = 4;
}else if (byg_start("PileUp",input[i]) != -1){
input_numseq[i] = count_sequences_clustalw(input[i]);
input_type[i] = 4;
}else if (byg_start("MSF:",input[i]) != -1){
input_numseq[i] = count_sequences_clustalw(input[i]);
input_type[i] = 4;
}else if (byg_start("STOCKHOLM",input[i]) != -1){
input_numseq[i] = count_sequences_stockholm(input[i]);
input_type[i] = 5;
}else{
input_numseq[i] = count_sequences_fasta(input[i]);
input_type[i] = 0;
}
k_printf("found %d sequences\n",input_numseq[i]);
if(input_numseq[i] < 1){
free(input[i]);
input[i] = 0;
}else{
numseq += input_numseq[i];
}
}else{
k_printf("found no sequences.\n");
if(!param->outfile && i){
param->outfile = param->infile[i];
k_printf("-> output file, in ");
//try to set format....
if(!param->format){
if (byg_start("msf",param->outfile) != -1){
param->format = "msf";
}else if (byg_start("clustal",param->outfile) != -1){
param->format = "clustal";
}else if (byg_start("aln",param->outfile) != -1){
param->format = "clustal";
}else if (byg_start("macsim",param->outfile) != -1){
param->format = "macsim";
}else{
param->format = "fasta";
}
if(param->reformat){
k_printf("unaligned fasta format\n");
}else if(param->format){
k_printf("%s format\n",param->format);
}else{
k_printf("fasta format\n");
}
}
}
k_printf("\n");
}
}
if(numseq < 2){
k_printf("%s\n", usage);
if(!numseq){
k_printf("\nWARNING: No sequences found.\n\n");
}else{
k_printf("\nWARNING: Only one sequence found.\n\n");
}
for (i = 0; i < num_input;i++){
free(input[i]);
}
free(input_numseq);
free(input_type);
free(input);
free_param(param);
exit(0);
}
if(byg_start(param->alignment_type,"profPROFprofilePROFILE") != -1){
if( free_read < 2){
k_printf("\nWARNING: You are trying to perform a profile - profile alignment but ony one input file was detected.\n\n");
param->alignment_type = "default";
}
}
if (param->feature_type && !feature){
for (i = 0; i < num_input;i++){
free(input[i]);
}
free(input_numseq);
free(input_type);
free(input);
free_param(param);
throwKalignException(k_printf("\nWARNING: You are trying to perform a feature alignment but the input format(s) do not contain feature information.\n"));
}
get_kalign_context()->numprofiles = (numseq << 1) - 1;
aln = aln_alloc(aln);
//numseq = 0;
if(byg_start(param->alignment_type,"profPROFprofilePROFILE") != -1){
j = 0;
for (i = 0; i < num_input;i++){
if(input[i]){
switch(input_type[i]){
case 0:
aln = read_alignment(aln,input[i]);
break;
case 1:
aln = read_alignment_macsim_xml(aln,input[i]);
break;
case 2:
aln = read_alignment_uniprot_xml(aln,input[i]);
break;
case 3:
aln = read_alignment_from_swissprot(aln, input[i]);
break;
case 4:
aln = read_alignment_clustal(aln,input[i]);
break;
case 5:
aln = read_alignment_stockholm(aln,input[i]);
break;
default:
aln = read_alignment(aln,input[i]);
break;
}
input[i] = 0;
//create partial profile....
aln->nsip[numseq+j] = input_numseq[i];
aln->sip[numseq+j] = malloc(sizeof(int)*aln->nsip[numseq+j]);
//k_printf("%d %d\n",numseq+j,aln->sl[numseq+j]);
j++;
}
}
num_input = j;
c = 0;
for (i = 0;i < num_input;i++){
//
for ( j = 0; j < aln->nsip[numseq+i];j++){
aln->sip[numseq+i][j] = c;
c++;
// k_printf("%d ",aln->sip[numseq+i][j]);
}
aln->sl[numseq+i] = aln->sl[aln->sip[numseq+i][0]];
// k_printf("PROFILE:%d contains: %d long:%d\n",i+numseq,aln->nsip[numseq+i],aln->sl[numseq+i]);
// k_printf("\n");
}
//sanity check -are all input
for (i = 0;i < num_input;i++){
for ( j = 0; j < aln->nsip[numseq+i]-1;j++){
a = aln->sip[numseq+i][j];
a = aln->sl[a];
for (c = j+1; j < aln->nsip[numseq+i];j++){
b = aln->sip[numseq+i][c];
b = aln->sl[b];
if(a != b){
for (i = 0; i < num_input;i++){
free(input[i]);
}
free(input_numseq);
free(input_type);
free(input);
free_aln(aln);
free_param(param);
throwKalignException(k_printf("Unaligned sequences in input %s.\n",param->infile[i]));
}
}
}
}
//exit(0);
/*for (i = 0; i < numseq;i++){
k_printf("len%d:%d\n",i,aln->sl[i]);
for ( j =0 ; j < aln->sl[i];j++){
//if(aln->s[i][j]> 23 || aln->s[i][j] < 0){
// aln->s[i][j] = -1;
//}
k_printf("%d ",aln->s[i][j]);
}
// k_printf("\n");
}
exit(0);*/
}else{
for (i = 0; i < num_input;i++){
if(input[i]){
switch(input_type[i]){
case 0:
aln = read_sequences(aln,input[i]);
break;
case 1:
aln = read_sequences_macsim_xml(aln,input[i]);
break;
case 2:
aln = read_sequences_uniprot_xml(aln,input[i]);
break;
case 3:
aln = read_sequences_from_swissprot(aln, input[i]);
break;
case 4:
aln = read_sequences_clustal(aln,input[i]);
break;
case 5:
aln = read_sequences_stockholm(aln,input[i]);
break;
default:
aln = read_sequences(aln,input[i]);
break;
}
/*if (byg_start("<macsim>",input[i]) != -1){
aln = read_sequences_macsim_xml(aln,input[i]);
}else if (byg_start("<uniprot",input[i]) != -1){
aln = read_sequences_uniprot_xml(aln,input[i]);
}else if(byg_start("This SWISS-PROT entry is copyright.",input[i]) != -1){
aln = read_sequences_from_swissprot(aln, input[i]);
}else if (byg_start("This Swiss-Prot entry is copyright.",input[i]) != -1){
aln = read_sequences_from_swissprot(aln, input[i]);
}else if (byg_start("CLUSTAL W",input[i]) != -1){
aln = read_sequences_clustal(aln,input[i]);
}else if (byg_start("PileUp",input[i]) != -1){
aln = read_sequences_clustal(aln,input[i]);
}else if (byg_start("MSF:",input[i]) != -1){
aln = read_sequences_clustal(aln,input[i]);
}else if (byg_start("STOCKHOLM",input[i]) != -1){
aln = read_sequences_stockholm(aln,input[i]);
}else{
aln = read_sequences(aln,input[i]);
}*/
input[i] = 0;
}
}
}
if(numseq < 2){
free_param(param);
throwKalignException(k_printf("\nNo sequences could be read.\n"));
}
if(!param->format && param->outfile){
if (byg_start("msf",param->outfile) != -1){
param->format = "msf";
}else if (byg_start("clustal",param->outfile) != -1){
param->format = "clustal";
}else if (byg_start("aln",param->outfile) != -1){
param->format = "clustal";
}else if (byg_start("macsim",param->outfile) != -1){
param->format = "macsim";
}
k_printf("Output file: %s, in %s format.\n",param->outfile,param->format);
}
free(input);
free(input_type);
free(input_numseq);
return aln;
}
void KalignAdapter::alignUnsafe(const MultipleSequenceAlignment& ma, MultipleSequenceAlignment& res, TaskStateInfo& ti) {
ti.progress = 0;
int* tree = 0;
quint32 a, b, c;
struct alignment* aln = 0;
struct parameters* param = 0;
struct aln_tree_node* tree2 = 0;
param = (parameters*)malloc(sizeof(struct parameters));
param = interface(param,0,0);
kalign_context *ctx = get_kalign_context();
unsigned int &numseq = ctx->numseq;
unsigned int &numprofiles = ctx->numprofiles;
if (ma->getNumRows() < 2){
if (!numseq){
k_printf("No sequences found.\n\n");
} else {
k_printf("Only one sequence found.\n\n");
}
free_param(param);
throw KalignException("Can't align less then 2 sequences");
}
if(ctx->gpo != -1) {
param->gpo = ctx->gpo;
}
if(ctx->gpe != -1) {
param->gpe = ctx->gpe;
}
if(ctx->tgpe != -1) {
param->tgpe = ctx->tgpe;
}
if(ctx->secret != -1) {
param->secret = ctx->secret;
}
/************************************************************************/
/* Convert MA to aln */
/************************************************************************/
k_printf("Prepare data");
numseq = ma->getNumRows();
numprofiles = (numseq << 1) - 1;
aln = aln_alloc(aln);
for(quint32 i = 0 ; i < numseq; i++) {
const MultipleSequenceAlignmentRow row= ma->getMsaRow(i);
aln->sl[i] = row->getUngappedLength();
aln->lsn[i] = row->getName().length();
}
for (quint32 i = 0; i < numseq;i++) {
try {
aln->s[i] = (int*) malloc(sizeof(int)*(aln->sl[i]+1));
checkAllocatedMemory(aln->s[i]);
aln->seq[i] = (char*)malloc(sizeof(char)*(aln->sl[i]+1));
checkAllocatedMemory(aln->seq[i]);
aln->sn[i] = (char*)malloc(sizeof(char)*(aln->lsn[i]+1));
checkAllocatedMemory(aln->sn[i]);
} catch (...) {
cleanupMemory(NULL, numseq, NULL, aln, param);
throw;
}
}
int aacode[26] = {0,1,2,3,4,5,6,7,8,-1,9,10,11,12,23,13,14,15,16,17,17,18,19,20,21,22};
for(quint32 i = 0; i < numseq; i++) {
const MultipleSequenceAlignmentRow row= ma->getMsaRow(i);
qstrncpy(aln->sn[i], row->getName().toLatin1(), row->getName().length() + 1); //+1 to include '\0'
QString gapless = QString(row->getCore()).remove('-');
qstrncpy(aln->seq[i], gapless.toLatin1(), gapless.length() + 1); //+1 to include '\0'
for (quint32 j = 0; j < aln->sl[i]; j++) {
if (isalpha((int)aln->seq[i][j])){
aln->s[i][j] = aacode[toupper(aln->seq[i][j])-65];
} else {
aln->s[i][j] = -1;
}
}
aln->s[i][aln->sl[i]] = 0;
aln->seq[i][aln->sl[i]] = 0;
aln->sn[i][aln->lsn[i]] = 0;
}
/*for(int i=0;i<numseq;i++) {
for(int j=0;j<aln->sl[i];j++)
printf("%d ", aln->s[i][j]);
}*/
//aln_dump(aln);
//aln = detect_and_read_sequences(aln,param);
if(param->ntree > (int)numseq){
param->ntree = (int)numseq;
}
//DETECT DNA
if(param->dna == -1){
for (quint32 i = 0; i < numseq;i++){
param->dna = byg_detect(aln->s[i],aln->sl[i]);
if(param->dna){
break;
}
}
}
//param->dna = 0;
//k_printf("DNA:%d\n",param->dna);
//exit(0);
if(param->dna == 1){
//brief sanity check...
for (quint32 i = 0; i < numseq;i++){
if(aln->sl[i] < 6){
//k_printf("Dna/Rna alignments are only supported for sequences longer than 6.");
free(param);
free_aln(aln);
throw KalignException("Dna/Rna alignments are only supported for sequences longer than 6.");
}
}
aln = make_dna(aln);
}
//int j;
//fast distance calculation;
float** submatrix = 0;
submatrix = read_matrix(submatrix,param); // sets gap penalties as well.....
//if(byg_start(param->alignment_type,"profPROFprofilePROFILE") != -1){
// profile_alignment_main(aln,param,submatrix);
//}
float** dm = 0;
if(param->ntree > 1){
//if(byg_start(param->distance,"pairclustalPAIRCLUSTAL") != -1){
// if(byg_start(param->tree,"njNJ") != -1){
// dm = protein_pairwise_alignment_distance(aln,dm,param,submatrix,1);
// }else{
// dm = protein_pairwise_alignment_distance(aln,dm,param,submatrix,0);
// }
//}else if(byg_start("wu",param->alignment_type) != -1){
// dm = protein_wu_distance2(aln,dm,param);
// // param->feature_type = "wumanber";
if(param->dna == 1){
// if(byg_start(param->tree,"njNJ") != -1){
// dm = dna_distance(aln,dm,param,1);
// }else{
dm = dna_distance(aln,dm,param,0);
// }
}else{
//if(byg_start(param->tree,"njNJ") != -1){
// dm = protein_wu_distance(aln,dm,param,1);
//}else{
try {
dm = protein_wu_distance(aln,dm,param,0);
} catch (const KalignException &) {
cleanupMemory(submatrix, numseq, dm, aln, param);
throw;
}
//}
}
if(check_task_canceled(ctx)) {
cleanupMemory(submatrix, numseq, dm, aln, param);
throwCancellingException();
}
/*int j;
for (int i = 0; i< numseq;i++){
for (j = 0; j< numseq;j++){
k_printf("%f ",dm[i][j]);
}
k_printf("\n");
}*/
//if(byg_start(param->tree,"njNJ") != -1){
// tree2 = real_nj(dm,param->ntree);
//}else{
tree2 = real_upgma(dm,param->ntree);
//}
//if(param->print_tree){
// print_tree(tree2,aln,param->print_tree);
//}
}
tree = (int*) malloc(sizeof(int)*(numseq*3+1));
for (quint32 i = 1; i < (numseq*3)+1;i++){
tree[i] = 0;
}
tree[0] = 1;
if(param->ntree < 2){
tree[0] = 0;
tree[1] = 1;
c = numseq;
tree[2] = c;
a = 2;
for (quint32 i = 3; i < (numseq-1)*3;i+=3){
tree[i] = c;
tree[i+1] = a;
c++;
tree[i+2] = c;
a++;
}
}else if(param->ntree > 2){
ntreeify(tree2,param->ntree);
}else{
tree = readtree(tree2,tree);
for (quint32 i = 0; i < (numseq*3);i++){
tree[i] = tree[i+1];
}
free(tree2->links);
free(tree2->internal_lables);
free(tree2);
}
//get matrices...
//struct feature_matrix* fm = 0;
//struct ntree_data* ntree_data = 0;
int** map = 0;
//if(param->ntree > 2){
// ntree_data = (struct ntree_data*)malloc(sizeof(struct ntree_data));
// ntree_data->realtree = tree2;
// ntree_data->aln = aln;
// ntree_data->profile = 0;
// ntree_data->map = 0;
// ntree_data->ntree = param->ntree;
// ntree_data->submatrix = submatrix;
// ntree_data->tree = tree;
// ntree_data = ntree_alignment(ntree_data);
// map = ntree_data->map;
// tree = ntree_data->tree;
// for (int i = 0; i < (numseq*3);i++){
// tree[i] = tree[i+1];
// }
// free(ntree_data);
//}else if (param->feature_type){
// fm = get_feature_matrix(fm,aln,param);
// if(!fm){
// for (int i = 32;i--;){
// free(submatrix[i]);
// }
// free(submatrix);
// free_param(param);
// free(map);
// free(tree);
// throw KalignException("getting feature matrix error");
// }
// map = feature_hirschberg_alignment(aln,tree,submatrix,map,fm);
// //exit(0);
// //map = feature_alignment(aln,tree,submatrix, map,fm);
//}else if (byg_start("pairwise",param->alignment_type) != -1){
// if(param->dna == 1){
// map = dna_alignment_against_a(aln,tree,submatrix, map,param->gap_inc);
// }else{
// map = hirschberg_alignment_against_a(aln,tree,submatrix, map,param->smooth_window,param->gap_inc);
// }
// //map = default_alignment(aln,tree,submatrix, map);
//}else if (byg_start("fast",param->alignment_type) != -1){
// map = default_alignment(aln,tree,submatrix, map);
if(param->dna == 1){
map = dna_alignment(aln,tree,submatrix, map,param->gap_inc);
// /*}else if (byg_start("test",param->alignment_type) != -1){
// map = test_alignment(aln,tree,submatrix, map,param->internal_gap_weight,param->smooth_window,param->gap_inc);
// }else if (param->aa){
// map = aa_alignment(aln,tree,submatrix, map,param->aa);
// }else if (param->alter_gaps){
// map = alter_gaps_alignment(aln,tree,submatrix,map,param->alter_gaps,param->alter_range,param->alter_weight);
// }else if (byg_start("altergaps",param->alignment_type) != -1){
// map = alter_gaps_alignment(aln,tree,submatrix,map,param->alter_gaps,param->alter_range,param->alter_weight);
// }else if(byg_start("simple",param->alignment_type) != -1){
// map = simple_hirschberg_alignment(aln,tree,submatrix, map);*/
//}else if(byg_start("advanced",param->alignment_type) != -1){
// map = advanced_hirschberg_alignment(aln,tree,submatrix, map,param->smooth_window,param->gap_inc,param->internal_gap_weight);
}else{
map = hirschberg_alignment(aln,tree,submatrix, map,param->smooth_window,param->gap_inc);
}
if (map == NULL) {
throw KalignException("Failed to build alignment.");
}
if(check_task_canceled(ctx)) {
free_param(param);
free_aln(aln);
free(map);
free(tree);
throwCancellingException();
}
//clear up sequence array to be reused as gap array....
int *p = 0;
for (quint32 i = 0; i < numseq;i++){
p = aln->s[i];
for (a = 0; a < aln->sl[i];a++){
p[a] = 0;
}
}
//clear up
for (quint32 i = 0; i < (numseq-1)*3;i +=3){
a = tree[i];
b = tree[i+1];
aln = make_seq(aln,a,b,map[tree[i+2]]);
}
//for (int i = 0; i < numseq;i++){
// k_printf("%s %d\n",aln->sn[i],aln->nsip[i]);
//}
for (quint32 i = 0; i < numseq;i++){
aln->nsip[i] = 0;
}
aln = sort_sequences(aln,tree,param->sort);
//for (int i = 0; i < numseq;i++){
// k_printf("%d %d %d\n",i,aln->nsip[i],aln->sip[i][0]);
//}
/************************************************************************/
/* Convert aln to MA */
/************************************************************************/
res->setAlphabet(ma->getAlphabet());
for (quint32 i = 0; i < numseq;i++){
int f = aln->nsip[i];
QString seq;
for(quint32 j=0;j<aln->sl[f];j++) {
seq += QString(aln->s[f][j],'-') + aln->seq[f][j];
}
seq += QString(aln->s[f][aln->sl[f]],'-');
res->addRow(QString(aln->sn[f]), seq.toLatin1());
}
//output(aln,param);
/* if(!param->format){
fasta_output(aln,param->outfile);
}else{
if (byg_start("msf",param->format) != -1){
msf_output(aln,param->outfile);
}else if (byg_start("clustal",param->format) != -1){
clustal_output(aln,param->outfile);
}else if (byg_start("macsim",param->format) != -1){
macsim_output(aln,param->outfile,param->infile[0]);
}
}
*/
free_param(param);
free_aln(aln);
free(map);
free(tree);
//KalignContext* ctx = getKalignContext();
}