SmithWaterman returns incorrect alignments.

[d-cameron]

The S/W aligner does not return optimal alignments.
Tested version: 0.8.6
Repo test case:

    @Test
    public void should_return_maximal_alignment_match() {
        IntelSmithWaterman isw = new IntelSmithWaterman();
        assertTrue(isw.load(null));
        SWNativeAlignerResult alignment = isw.align(
                "GGGGGGTTTTTT".getBytes(),
                "AAACCCTTTTTT".getBytes(),
                // BWA alignment scoring matrix
                new SWParameters(1, -4, -6, -1),
                SWOverhangStrategy.SOFTCLIP);
        assertEquals("6S6M", alignment.cigar);
        assertEquals(6, alignment.alignment_offset);
    }

Actual alignment cigar returned is 11S1M. This is clearly a sub-optimal alignment.

I strongly recommend adding a test suite to exhaustively compare small results against another S/W aligner (e.g. JAligner) to ensure correctness of your implementation.

[lbergelson]

That's worrisome. We've never enabled the optimized smith waterman by default in GATK because of some other issues, but I hadn't ever seen it produce grossly incorrect alignments. I was under the impression that it had been tested pretty heavily against the GATK java implementation. I wonder if we produce the same bad alignment...

[lbergelson]

I'm confused about this one. Running this code against the GATK java smith waterman implementation produces the same result. It seems to have something to do with the overhang strategy although none of the overhang strategies produce the answer that seems most correct. @davidbenjamin do you have any sense of what's going on here?

[davidbenjamin]

@lbergelson See comments in broadinstitute/gatk#6576.

[Kmannth]

Still blocked on broadinstitute/gatk#6576