int main(int argc, char *argv[]) {
FILE *fileHandle;
struct List *seqs;
struct List *seqLengths;
struct List *seqNames;
int32_t i;
assert(argc == 3);
seqs = constructEmptyList(0, free);
seqLengths = constructEmptyList(0, (void (*)(void *))destructInt);
seqNames = constructEmptyList(0, free);
fileHandle = fopen(argv[1], "r");
fastaRead(fileHandle, seqs, seqLengths, seqNames);
fclose(fileHandle);
fileHandle = fopen(argv[2], "w");
for(i=0; i < seqs->length; i++) {
assert(strlen(seqs->list[i]) == *((int32_t *)seqLengths->list[i]));
fastaWrite(seqs->list[i], seqNames->list[i], fileHandle);
}
fclose(fileHandle);
return 0;
}
// Returns a hash mapping from sequence header to sequence data.
static stHash *readFastaFile(char *filename) {
FILE *fasta = fopen(filename, "r");
if (fasta == NULL) {
st_errnoAbort("Could not open fasta file %s", filename);
}
stHash *headerToData = stHash_construct3(stHash_stringKey,
stHash_stringEqualKey,
free,
free);
struct List *seqs = constructEmptyList(0, NULL);
struct List *seqLengths = constructEmptyList(0, free);
struct List *headers = constructEmptyList(0, free);
fastaRead(fasta, seqs, seqLengths, headers);
for (int64_t i = 0; i < seqs->length; i++) {
char *fullHeader = headers->list[i];
stList *headerTokens = stString_splitByString(fullHeader, " ");
char *usableHeader = stString_copy(stList_get(headerTokens, 0));
stHash_insert(headerToData, usableHeader, seqs->list[i]);
stList_destruct(headerTokens);
}
destructList(seqs);
destructList(seqLengths);
destructList(headers);
return headerToData;
}
Event *eventTree_getCommonAncestor(Event *event, Event *event2) {
Event *ancestorEvent;
struct List *list;
assert(event != NULL);
assert(event2 != NULL);
assert(event_getEventTree(event) == event_getEventTree(event2));
list = constructEmptyList(0, NULL);
ancestorEvent = event;
while(ancestorEvent != NULL) {
if(ancestorEvent == event2) {
destructList(list);
return event2;
}
listAppend(list, ancestorEvent);
ancestorEvent = event_getParent(ancestorEvent);
}
ancestorEvent = event2;
while((ancestorEvent = event_getParent(ancestorEvent)) != NULL) {
if(listContains(list, ancestorEvent)) {
destructList(list);
return ancestorEvent;
}
}
destructList(list);
assert(FALSE);
return NULL;
}
static stList *getRandomPairwiseAlignments() {
stList *pairwiseAlignments = stList_construct3(0, (void(*)(void *)) destructPairwiseAlignment);
int64_t randomAlignmentNumber = st_randomInt(0, 10);
for (int64_t i = 0; i < randomAlignmentNumber; i++) {
char *contig1 = stString_print("%" PRIi64 "", i);
char *contig2 = stString_print("%" PRIi64 "", i * 10);
int64_t start1 = st_randomInt(100000, 1000000);
int64_t start2 = st_randomInt(100000, 1000000);
int64_t strand1 = st_random() > 0.5;
int64_t strand2 = st_random() > 0.5;
int64_t end1 = start1;
int64_t end2 = start2;
struct List *operationList = constructEmptyList(0, NULL);
while (st_random() > 0.1) {
int64_t length = st_randomInt(0, 10);
int64_t type = st_randomInt(0, 3);
assert(type < 3);
listAppend(operationList, constructAlignmentOperation(type, length, 0));
if (type != PAIRWISE_INDEL_Y) {
end1 += strand1 ? length : -length;
}
if (type != PAIRWISE_INDEL_X) {
end2 += strand2 ? length : -length;
}
}
stList_append(pairwiseAlignments,
constructPairwiseAlignment(contig1, start1, end1, strand1, contig2, start2, end2, strand2, 0.0, operationList));
free(contig1);
free(contig2);
}
return pairwiseAlignments;
}
struct List *parseTrioFile(char *trioFile) {
FILE *fileHandle = fopen(trioFile, "r");
int bytesRead;
int nBytes = 100;
char *cA;
int j;
char *species[3];
struct List *speciesList = NULL;
speciesList = constructEmptyList(0, freeTrioNames);
cA = st_malloc(nBytes + 1);
bytesRead = benLine(&cA, &nBytes, fileHandle);
while(bytesRead != -1) {
if (bytesRead > 0) {
species[0] = st_malloc(sizeof(char) * (1 + (bytesRead)));
species[1] = st_malloc(sizeof(char) * (1 + (bytesRead)));
species[2] = st_malloc(sizeof(char) * (1 + (bytesRead)));
j = sscanf(cA, "%s\t%s\t%s", species[0], species[1], species[2]);
if (j != 3) {
fprintf(stderr, "Invalid triple line '%s' in '%s'\n", cA, trioFile);
exit(1);
}
cStr_lowerCase(species[0]);
cStr_lowerCase(species[1]);
cStr_lowerCase(species[2]);
qsort(species, 3, sizeof(char *), cStr_compare);
TrioNames *trio = st_malloc(sizeof(TrioNames));
trio->speciesA = species[0];
trio->speciesB = species[1];
trio->speciesC = species[2];
listAppend(speciesList, trio);
}
bytesRead = benLine(&cA, &nBytes, fileHandle);
}
fclose(fileHandle);
free(cA);
return speciesList;
}
Event *event_construct(Name name, const char *header, float branchLength, Event *parentEvent,
EventTree *eventTree) {
assert(eventTree_getEvent(eventTree, name) == NULL); //the event must not already exist in the tree.
Event *event;
event = st_malloc(sizeof(Event));
event->name = name;
event->parent = parentEvent;
event->children = constructEmptyList(0, NULL);
event->header = stString_copy(header == NULL ? "" : header);
event->branchLength = branchLength < 0.0 ? 0.0 : branchLength;
event->isOutgroup = 0;
if (parentEvent != NULL) {
listAppend(parentEvent->children, event);
}
event->eventTree = eventTree;
eventTree_addEvent(eventTree, event);
return event;
}
// copied from cPecanRealign
struct PairwiseAlignment *convertAlignedPairsToPairwiseAlignment(char *seqName1, char *seqName2, double score,
int64_t length1, int64_t length2, stList *alignedPairs) {
//Make pairwise alignment
int64_t pX = -1, pY = -1, mL = 0;
//Create an end matched pair, which is used to ensure the alignment has the correct end indels.
struct List *opList = constructEmptyList(0, (void (*)(void *)) destructAlignmentOperation);
stList_append(alignedPairs, stIntTuple_construct2(length1, length2));
for (int64_t i = 0; i < stList_length(alignedPairs); i++) {
stIntTuple *alignedPair = stList_get(alignedPairs, i);
int64_t x = stIntTuple_get(alignedPair, 0);
int64_t y = stIntTuple_get(alignedPair, 1);
assert(x - pX > 0);
assert(y - pY > 0);
if (x - pX > 0 && y - pY > 0) { //This is a hack for filtering
if (x - pX > 1) { //There is an indel.
if (mL > 0) {
listAppend(opList, constructAlignmentOperation(PAIRWISE_MATCH, mL, 0));
mL = 0;
}
listAppend(opList, constructAlignmentOperation(PAIRWISE_INDEL_X, x - pX - 1, 0));
}
if (y - pY > 1) {
if (mL > 0) {
listAppend(opList, constructAlignmentOperation(PAIRWISE_MATCH, mL, 0));
mL = 0;
}
listAppend(opList, constructAlignmentOperation(PAIRWISE_INDEL_Y, y - pY - 1, 0));
}
mL++;
pX = x;
pY = y;
}
}
//Deal with a trailing match, but exclude the final match
if (mL > 1) {
listAppend(opList, constructAlignmentOperation(PAIRWISE_MATCH, mL - 1, 0));
}
stIntTuple_destruct(stList_pop(alignedPairs));
//Construct the alignment
struct PairwiseAlignment *pA = constructPairwiseAlignment(seqName1, 0, length1, 1, seqName2, 0, length2, 1, score,
opList);
return pA;
}
int main(int argc, char *argv[]) {
/*
* Arguments/options
*/
char *logLevelString = NULL;
char *mfaFile = NULL;
char *outputFile = NULL;
char *treeFile = NULL;
///////////////////////////////////////////////////////////////////////////
// (0) Parse the inputs handed by genomeCactus.py / setup stuff.
///////////////////////////////////////////////////////////////////////////
while (1) {
static struct option long_options[] = {
{ "logLevel", required_argument, 0, 'a' },
{ "mfaFile", required_argument, 0, 'b' },
{ "outputFile", required_argument, 0, 'd' },
{ "treeFile", optional_argument, 0, 't' },
{ "help", no_argument, 0, 'h' },
{ 0, 0, 0, 0 }
};
int option_index = 0;
int key = getopt_long(argc, argv, "a:b:d:t:h", long_options, &option_index);
if (key == -1) {
break;
}
switch (key) {
case 'a':
logLevelString = stString_copy(optarg);
break;
case 'b':
mfaFile = stString_copy(optarg);
break;
case 'd':
outputFile = stString_copy(optarg);
break;
case 't':
treeFile = stString_copy(optarg);
break;
case 'h':
usage();
return 0;
default:
usage();
return 1;
}
}
///////////////////////////////////////////////////////////////////////////
// (0) Check the inputs.
///////////////////////////////////////////////////////////////////////////
if (argc == 1) {
usage();
exit(1);
}
assert(mfaFile != NULL);
assert(outputFile != NULL);
//////////////////////////////////////////////
//Set up logging
//////////////////////////////////////////////
st_setLogLevelFromString(logLevelString);
//////////////////////////////////////////////
//Log (some of) the inputs
//////////////////////////////////////////////
st_logInfo("MFA file name : %s\n", mfaFile);
st_logInfo("Output MAF file : %s\n", outputFile);
st_logInfo("Tree file name: %s\n", treeFile == NULL ? "null" : treeFile);
//////////////////////////////////////////////
//Get the MFA alignment
//////////////////////////////////////////////
//get the alignment
struct List *sequences = constructEmptyList(0, free);
struct List *seqLengths = constructEmptyList(0, (void (*)(void *))destructInt);
struct List *fastaNames = constructEmptyList(0, free);
FILE *fileHandle = fopen(mfaFile, "r");
if (fileHandle == NULL) {
usage();
exit(1);
}
fastaRead(fileHandle, sequences, seqLengths, fastaNames);
fclose(fileHandle);
//////////////////////////////////////////////
//Get the tree alignment
//////////////////////////////////////////////
stTree *tree = NULL;
LeafPtrArray *leafArray = NULL;
int32_t leafCount = 0;
if (treeFile != NULL) {
tree = eTreeX_getTreeFromFile(treeFile);
eTreeX_postOrderTraversal(tree, eTreeX_countLeaves, &leafCount);
leafArray = eTreeX_constructLeafPtrArray(leafCount);
eTreeX_postOrderTraversal(tree, eTreeX_getLeafArray, (void *) leafArray);
}
//////////////////////////////////////////////
//Write the MFA alignment.
//////////////////////////////////////////////
fileHandle = fopen(outputFile, "w");
//write the header.
fprintf(fileHandle, "##maf version=1 scoring=NULL\n");
fprintf(fileHandle, "# converted_from_MFA\n\n");
//write the score line
char *treeString = NULL;
if (treeFile != NULL) {
treeString = stTree_getNewickTreeString(tree);
fprintf(fileHandle, "a score=0 tree=\"%s\"\n", treeString);
}
else {
fprintf(fileHandle, "a score=0\n");
leafCount = sequences->length;
}
//write the alignment
int32_t i, j;
int32_t ii;
const char *label;
for (ii=0; ii<leafCount; ii++) {
if (treeFile != NULL) {
label = stTree_getLabel((stTree *) leafArray->ptrArray[ii]);
/* Do a brute force search to find the appropriate sequence that matches "label" */
for (i=0; i<sequences->length; i++) {
char *fastaHeader = fastaNames->list[i];
char *sequenceName = st_malloc(sizeof(char) *(1 + strlen(fastaHeader)));
sscanf(fastaHeader, "%s", sequenceName); //take the sequence name to be the first word of the sequence.
if (strcmp(label, sequenceName) == 0) {
free(sequenceName);
break;
}
free(sequenceName);
}
}
else {
i = ii;
}
char *sequence = sequences->list[i];
int32_t seqLength = *((int32_t *)seqLengths->list[i]);
assert(seqLength == (int32_t)strlen(sequence));
char *fastaHeader = fastaNames->list[i];
char *sequenceName = st_malloc(sizeof(char) *(1 + strlen(fastaHeader)));
sscanf(fastaHeader, "%s", sequenceName); //take the sequence name to be the first word of the sequence.
int32_t length = 0;
for (j=0; j<(int32_t)strlen(sequence); j++) {
if (sequence[j] != '-') {
length++;
}
}
fprintf(fileHandle, "s\t%s\t%i\t%i\t%s\t%i\t%s\n", sequenceName, 0, length, "+", length, sequence);
free(sequenceName);
}
fclose(fileHandle);
//////////////////////////////////////////////
//Clean up.
//////////////////////////////////////////////
free(mfaFile);
free(outputFile);
free(treeFile);
if (treeFile != NULL) {
stTree_destruct(tree);
free(treeString);
eTreeX_destructLeafPtrArray(leafArray);
}
destructList(sequences);
destructList(seqLengths);
destructList(fastaNames);
return 0;
}
