TEST_F(VectorMethods, ReverseCompletes) {
Vector reversed = {};
SystemErr eIgnore = S_E_CLEAR;
initVector(&reversed, sizeof(int), NULL, NULL, &eIgnore);
int nums[3] = { 1, 2, 3 };
Vector_catPrimitive(&v, nums, 3, &eIgnore);
Vector_reverse(&v, &reversed, &eIgnore);
EXPECT_EQ(*(int*) reversed.arr, 3);
EXPECT_EQ(*((int*) reversed.arr + 1), 2);
EXPECT_EQ(*((int*) reversed.arr + 2), 1);
deinitVector(&reversed);
}
void GenomicAlignAdaptor_addDerivedAlignments(GenomicAlignAdaptor *gaa,
Vector *mergedAligns, GenomicAlign *alignA, GenomicAlign *alignB) {
// variable name explanation
// q - query c - consensus s - start e - end l - last
// o, ov overlap j - jump_in_
// r - result
int qs, qe, lqs, lqe, cs, ce, lce,
ocs, oce, oqs, oqe, jc, jq, ovs, ove,
rcs, rce, rqs, rqe;
int currentMatch = 0;
int newMatch;
int cigAPos = 0, cigBPos = 0;
char *resultCig;
char tmpStr[128];
// initialization phase
Vector *cigA = CigarStrUtil_getPieces(GenomicAlign_getCigarString(alignA));
Vector *cigB = CigarStrUtil_getPieces(GenomicAlign_getCigarString(alignB));
if (GenomicAlign_getQueryStrand(alignA) == -1 ) {
Vector_reverse(cigB);
}
// need a 'normalized' start for qs, qe, oxs so I dont
// have to check strandedness all the time
// consensus is strand 1 and is not compared to anything,
// can keep its original coordinate system
lce = GenomicAlign_getConsensusStart(alignA) - 1;
ce = lce;
cs = ce + 1;
// alignBs query can be + or - just keep relative coords for now
lqe = 0; lqs = 1;
qe = 0; qs = 1;
// ocs will be found relative to oce and has to be comparable
// to oqs. But it could be that we have to move downwards if we
// are not - strand. thats why coordinates are transformed here
if (GenomicAlign_getQueryStrand(alignA) == -1 ) {
// query_end is first basepair of alignment
if (GenomicAlign_getQueryEnd(alignA) < GenomicAlign_getConsensusEnd(alignB)) {
oce = 0; ocs = 1;
oqe = GenomicAlign_getConsensusEnd(alignB) - GenomicAlign_getQueryEnd(alignA);
oqs = oqe + 1;
} else {
oqe = 0; oqs = 1;
oce = GenomicAlign_getQueryEnd(alignA) - GenomicAlign_getConsensusEnd(alignB);
ocs = oce + 1;
}
} else {
// in theory no coordinate magic necessary :-)
oqs = GenomicAlign_getQueryStart(alignA);
oqe = oqs - 1;
ocs = GenomicAlign_getConsensusStart(alignB);
oce = ocs - 1;
}
// initializing result
rcs = rce = rqs = rqe = 0;
resultCig= StrUtil_copyString(&resultCig,"",0);
while (1) {
int newGa;
// exit if you request a new piece of alignment and the cig list is
// empty
if (oce < ocs || oce < oqs) {
// next M area in cigB
if (cigBPos == Vector_getNumElement(cigB)) break;
GenomicAlignAdaptor_nextCig(gaa, cigB, &cigBPos, &ocs, &oce, &qs, &qe );
continue;
}
if (oqe < oqs || oqe < ocs) {
// next M area in cigA
if (cigAPos == Vector_getNumElement(cigA)) break;
GenomicAlignAdaptor_nextCig(gaa, cigA, &cigAPos, &cs, &ce, &oqs, &oqe );
continue;
}
// now matching region overlap in reference genome
ovs = ocs < oqs ? oqs : ocs;
ove = oce < oqe ? oce : oqe;
if (currentMatch) {
jc = cs + (ovs - oqs) - lce - 1;
jq = qs + (ovs - ocs) - lqe - 1;
} else {
jc = jq = 0;
}
newMatch = ove - ovs + 1;
newGa = 0;
if (jc==0) {
if (jq==0) {
currentMatch += newMatch;
} else {
// store current match;
sprintf(tmpStr,"%dM",currentMatch);
resultCig = StrUtil_appendString(resultCig,tmpStr);
// jq deletions;
if (jq == 1) {
resultCig = StrUtil_appendString(resultCig,"D");
} else {
sprintf(tmpStr,"%dD",jq);
resultCig = StrUtil_appendString(resultCig,tmpStr);
}
currentMatch = newMatch;
}
} else {
if (jq==0) {
// store current match;
sprintf(tmpStr,"%dM",currentMatch);
resultCig = StrUtil_appendString(resultCig,tmpStr);
// jc insertions;
if (jc==1) {
resultCig = StrUtil_appendString(resultCig,"I");
} else {
sprintf(tmpStr,"%dI",jc);
resultCig = StrUtil_appendString(resultCig,tmpStr);
}
currentMatch = newMatch;
} else {
double percId;
double score;
GenomicAlign *ga;
sprintf(tmpStr,"%dM",currentMatch);
resultCig = StrUtil_appendString(resultCig,tmpStr);
// new GA
int queryStrand = GenomicAlign_getQueryStrand(alignA) * GenomicAlign_getQueryStrand(alignB);
int queryStart, queryEnd;
if (queryStrand == 1) {
queryStart = rqs + GenomicAlign_getQueryStart(alignB) - 1;
queryEnd = rqe + GenomicAlign_getQueryStart(alignB) - 1;
} else {
queryEnd = GenomicAlign_getQueryEnd(alignB) - rqs + 1;
queryStart = GenomicAlign_getQueryEnd(alignB) - rqe + 1;
}
score = (GenomicAlign_getScore(alignA) < GenomicAlign_getScore(alignB)) ?
GenomicAlign_getScore(alignA) : GenomicAlign_getScore(alignB);
percId = (int)(GenomicAlign_getPercentId(alignA)*GenomicAlign_getPercentId(alignB)/100.0);
ga = GenomicAlign_new();
GenomicAlign_setConsensusDNAFrag(ga, GenomicAlign_getConsensusDNAFrag(alignA));
GenomicAlign_setQueryDNAFrag(ga, GenomicAlign_getQueryDNAFrag(alignB));
GenomicAlign_setCigarString(ga, resultCig);
GenomicAlign_setConsensusStart(ga, rcs);
GenomicAlign_setConsensusEnd(ga, rce);
GenomicAlign_setQueryStrand(ga, queryStrand);
GenomicAlign_setQueryStart(ga, queryStart);
GenomicAlign_setQueryEnd(ga, queryEnd);
GenomicAlign_setAdaptor(ga, (BaseAdaptor *)gaa);
GenomicAlign_setPercentId(ga, percId);
GenomicAlign_setScore(ga, score);
Vector_addElement(mergedAligns, ga);
rcs = rce = rqs = rqe = 0;
resultCig[0] = '\0';
currentMatch = newMatch;
}
}
if (!rcs) rcs = cs+(ovs-oqs);
rce = cs+(ove-oqs);
if (!rqs) rqs = qs+(ovs-ocs);
rqe = qs+(ove-ocs);
// update the last positions
lce = rce;
lqe = rqe;
// next piece on the one that end earlier
if (oce <= oqe) {
// next M area in cigB
if (cigBPos == Vector_getNumElement(cigB)) break;
GenomicAlignAdaptor_nextCig(gaa, cigB, &cigBPos, &ocs, &oce, &qs, &qe );
}
if (oce >= oqe) {
// next M area in cigA
if (cigAPos == Vector_getNumElement(cigA)) break;
GenomicAlignAdaptor_nextCig(gaa, cigA, &cigAPos, &cs, &ce, &oqs, &oqe );
}
} // end of while loop
// if there is a last floating current match
if (currentMatch) {
// new GA
int queryStrand = GenomicAlign_getQueryStrand(alignA) * GenomicAlign_getQueryStrand(alignB);
int queryStart, queryEnd;
double percId;
double score;
GenomicAlign *ga;
sprintf(tmpStr,"%dM",currentMatch);
resultCig = StrUtil_appendString(resultCig, tmpStr);
if (queryStrand == 1) {
queryStart = rqs + GenomicAlign_getQueryStart(alignB) - 1;
queryEnd = rqe + GenomicAlign_getQueryStart(alignB) - 1;
} else {
queryEnd = GenomicAlign_getQueryEnd(alignB) - rqs + 1;
queryStart = GenomicAlign_getQueryEnd(alignB) - rqe + 1;
}
score = (GenomicAlign_getScore(alignA) < GenomicAlign_getScore(alignB)) ?
GenomicAlign_getScore(alignA) : GenomicAlign_getScore(alignB);
percId = (int)(GenomicAlign_getPercentId(alignA)*GenomicAlign_getPercentId(alignB)/100.0);
ga = GenomicAlign_new();
GenomicAlign_setConsensusDNAFrag(ga, GenomicAlign_getConsensusDNAFrag(alignA));
GenomicAlign_setQueryDNAFrag(ga, GenomicAlign_getQueryDNAFrag(alignB));
GenomicAlign_setCigarString(ga, resultCig);
GenomicAlign_setConsensusStart(ga, rcs);
GenomicAlign_setConsensusEnd(ga, rce);
GenomicAlign_setQueryStrand(ga, queryStrand);
GenomicAlign_setQueryStart(ga, queryStart);
GenomicAlign_setQueryEnd(ga, queryEnd);
GenomicAlign_setAdaptor(ga, (BaseAdaptor *)gaa);
GenomicAlign_setPercentId(ga, percId);
GenomicAlign_setScore(ga, score);
Vector_addElement(mergedAligns, ga);
}
free(resultCig);
Vector_free(cigA);
Vector_free(cigB);
// nothing to return all in merged_aligns
}