Bowtie2 alignment scores seem less than the actual value

Question

[Sina]

As a test I am aligning the following reads

@sequence1
CCTCATCAGGCACCCTG
+
FFFFFFFFFFFFFFFFF
@sequence2-X
CCTCATCAGGCACCTCTG
+
FFFFFFFFFFFFFFFFFF

to the reference sequence

CCTCATCAGGCACCCTG

using the arguments (just to play around and understand with alignment scores and penalties)

--ma 2 --mp 1,1 --rdg 1,1 --rfg 1,1 --local --score-min C,10 -N 1 -L 2 -i C,1 --threads 12 -a

Three of the alignments I get are

sequence1   0   0   1   39  17M *   0   0   CCTCATCAGGCACCCTG   FFFFFFFFFFFFFFFFF   AS:i:34 XS:i:14 XN:i:0  XM:i:0  XO:i:0  XG:i:0  NM:i:0  MD:Z:17 YT:Z:UU

sequence2-X 272 0   1   25  8S10M   *   0   0   CAGAGGTGCCTGATGAGG  FFFFFFFFFFFFFFFFFF  AS:i:14 XS:i:14 XN:i:0  XM:i:2  XO:i:0  XG:i:0  NM:i:2  MD:Z:3C2C3  YT:Z:UU

sequence2-X 0   0   1   25  12M1I5M *   0   0   CCTCATCAGGCACCTCTG  FFFFFFFFFFFFFFFFFF  AS:i:29 XS:i:14 XN:i:0  XM:i:1  XO:i:1  XG:i:1  NM:i:2  MD:Z:13C3   YT:Z:UU

As you can see first one is 17M and so score is 17x2 = 34 as denoted in the AS field. But second alignment is 8S10M and since this is local should be 20 but alignment score is 14. Similarly third one is 12M1I5M which should be 17x2 - 1-1 = 32 but it denotes the score as 29. The second one is 6 lower and the third one 3. Am I miscalculating alignment scores?

Thanks

Answer 1

[Istvan Albert]

The M character of the CIGAR string denotes match or mismatch.

The number of mismatches NM:i:2 tag shows that the second alignment two has two mismatch penalties whereas the third has a mismatch and insertion penalty.

[Sina]

Huh, thanks I thought M was always a match should have looked at the sequences manually. Now that I did it is obvious there should have been some mismatches.