Dearest Armin,
for my NGS project I prefer using bwa for UDPS data. It would be fabulous of you if you could implement my wishlist:
Introduce a majority mode, respecting pluralityN. That is, if the fraction of gaps is below N, use the most frequent base for the consensus instead of all bases above the plurality. This is important for BWA, which has some animosity with regards to wobbles.
Possibly include insertions. If say some conserved insert appears at a position above (another given) threshold, integrate it too into the consensus.
Dearest Armin, for my NGS project I prefer using bwa for UDPS data. It would be fabulous of you if you could implement my wishlist:
Hugz and Kisses