Score alignments?

We are interested in the quality of an alignment of the query to the coordinate reference, because we want to quantify how certain we are about making calls e.g. in regard to drug resistance interpretations, and the coverage alone is not a good measure - we also want to know how well the assembled product actually matches the coordinate reference we use to identify regions and key positions in the genome.
Our first approach to this is to calculate and score the concordance, i.e. the amount of matching nucleotides within the alignment's matches combined with the amount of nucleotides that are covered by a match (see #821).
We also thought about additionally scoring alignments, to quantify how well the query matches the coordinate reference. For example, a large match should get a good score, whereas a fragmented assembly with a lot of indels should get a bad score.

To do:

• Research what other people do in this situation. Could we just use alignment scores?
• Implement the scoring system.
• Compare alignment scores across samples.