Futile gap opening

[jnvaughn]

Hi,

When using progressiveMauve (default parameters), we often observe an issue I've call "futile gap opening". Here is an example:

ACTAGCTGATG--------CTGACGTAATCGTGATGATCGATGCTGA 
ACTAGCTGATGCTGACGTA--------ATCGTGATGATCGATGCTGA
ACTAGCTGATGCTGACGTA--------ATCGTGATGATCGATGCTGA

As you can see, these gaps are clearly not warranted.

The issue is quite common. The histogram below is from whole genome alignments. I won't describe the x-axis in detail but the sharp peak at 0.5 reflect these futile gap openings as a fraction of all gap events in the total alignment.

Does anyone know the nature of this issue and how to avoid?

Thanks.