Score HLA coverage by nucleotides?

Because HLA's nucleotide sequence is interrupted by introns, we can't translate directly from the mapped nucleotide sequence to an amino acid sequence that will match the reference sequence. The boundaries of the introns are not on codon boundaries, so there is one codon split between exon1 and exon2, and there is another codon split between exon2 and exon3. It's possible that we could get a mutation in the intron that would appear to change the first codon of exon2, even though the end of exon1 hadn't changed. This might cause an incorrect drop in the coverage map.

Possible solutions:

• Score HLA coverage by nucleotides, that way the codon boundaries don't matter.
• Combine the exons together before translating to amino acids and reporting coverage.
• Do nothing, because we haven't seen any problems yet.

No files, tests, or entry points are named. Start by locating the HLA coverage and exon-translation logic, then compare nucleotide scoring with concatenating exons before translation. Done requires choosing and implementing one approach, with tests covering codons split across exon boundaries.