In particular, this defines the output SAM to be unsorted BUT also query grouped. The latter is very important to explicitly define so downstream tools that don't make assumptions know that reads from the same template are grouped.
In the bwa.c and bwase.c calls, rlen is an int64_t returned from
bns_get_seq() and is the number of reference bases covered by the
alignment; l_query/len is an int and the query length of the alignment;
and the result is an int given to an int parameter of ksw_global[2]().
As even the result is int and as rlen is effectively bounded by the
maximum length of a reference sequence, we maintain the status quo in
this code and simply cast rlen to int to silence Clang's "use llabs()"
(llabs() would not be a great answer given an int64_t anyway).
The bwtsw2_pair.c call needs to remain fabs() so both divisions are
done in floating point; cast to double to prevent Clang suggesting
changing the call to integer abs().
This function causes all kinds of problems when the reference genome consists
of many short reads/contigs/chromsomes. Some of the problems are nearly
unfixable at the point where bwa_fix_xref() gets called. This commit attempts
to fix the problem at the root. It disallows chains spanning multiple contigs
and never retrieves sequences bridging two adjacent contigs. Thus all the
chaining, extension, SW and global alignments are confined to on contig only.
This commit brings many changes. I have tested it on a couple examples
including Peter Field's PacBio example. It works well so far.
Peter Field has sent me an example caused by an alignment bridging three
adjacent chromosomes/contigs. Bwa-mem always aligns the query to the contig
covering the middle point of the alignment. In this example, it chooses the
middle contig, which should not be aligned. This leads to weird things failing
bwa_fix_xref2(), which cannot be fixed unless we build the contig boundaries
into the FM-index.
In the old code, bwa-mem halts when bwa_fix_xref2() fails. With this commit,
bwa-mem will give a warning instead of halting.
The old method does not work when the alignment bridges three chr. This may
actually happen often. The new method does not work all the time, either, but
should be better than the old one. It is also simpler, arguably.
zzh
zzh
Nils Homer
Heng Li
John Marshall
James Blachly