Compute percent identity using only the portion of the alignment used in the coverage calculation.

[jonathancrabtree]

Only the part of the alignment that's used to determine coverage should be used to determine percent identity. See #9.

[jonathancrabtree]

For example, the first two bases of the following ALU alignment would not be considered when computing the percent identity because they overlap with the TSD and are hence excluded from the coverage calculation:

chr22:25792353  |ALU  |+  |100.0%| 95.1%| 97.5%|100.0%| ATGCTGAGAT [AAGAAGTGGATTGGCCGGGCGCGGTGGCTCACGCCTGTAATCCCA....+224bp....CGACCTGGGCGACAGAGCGAGACTCCGTCTCAAAAAAAAAAAAAA] AAGAAGTGGATTGGTGAGATGCTGAGCGAC
                                                                    ^^^^^TSD^^^^^^                                                                            <---polyA---->  ^^^^^TSD^^^^^^
                                                                                [ALU-----------------------------              ---------------------------ALU>

[jonathancrabtree]

Probably not high priority because you can see in the current output which alignments have a significant difference between the gapped and ungapped percent identity.

[jonathancrabtree]

old "nogaps" percent identity = 90.4%, new= 77% (old "total" = 34%):