Have a look at the Cigar.java source code in Picard. For example:
/**
* @return The number of reference bases that the read covers, including padding.
*/
public int getPaddedReferenceLength() {
int length = 0;
for (final CigarElement element : cigarElements) {
switch (element.getOperator()) {
case M:
case D:
case N:
case EQ:
case X:
case P:
length += element.getLength();
}
}
return length;
}
•
written
by
Pierre Lindenbaum
Once you have length, the position of the last base is: end = start + length - 1. Don't forget the minus one, since we're "counting fenceposts" (google it).
The bamToBed program in BEDTools will do this by create BED entries for each BAM alignment where the end end coordinate reflects the CIGAR string. Moreover, if you want to create separate BED entries for "spliced" or "split" alignments (i.e., when there is an "N: CIGAR op present), use the -split option.
•
written
by
Aaronquinlan
Just remember, BED entries are 0-indexed, half-open ranges (up to but not including the end position). A 5 bp read starting at (1-based) position 10 would be "chr1 9 15" in BED.
Hi
I am trying the split option on my bam file and it is only reporting 6 fields in the output and that too for all the alignments. Can you suggest what is going wrong??
Log in to answer this question.
