DotPlotly appears to move data from one contig to another

[jbgonzalez77]

Hello,

I am excited to use Mummer/Nucmer, and your DotPlotly is great, however I think I may have found a bug. Before I scale up to bigger projects, I am running an alignment between a couple of small datasets and comparing this to the output from a Mauve alignment (same data).

My ref dataset has 3 contigs, my query dataset has 4. For some reason, in the DotPlotly output, a large portion of one of the contigs in the query dataset appears to have moved to a different contig. The query contigs (C4s12 and C4s13) are of nearly the same length, but in the DotPlotly visualization of the alignment, C4s12 is now much shorter, and the missing region appears to be attributed to C4s13.

This obviously has huge ramifications on the interpretation of genome reorganization events, which is what I am working on. Please see attached image showing both Mauve and Nucmer alignments (visualized using DotPlotly), with the region of interest circled in red. I have checked the Nucmer coords file, and this is problem is not there, the coords file shows correct attribution of alignments to contigs. I have have also run this same visualization using Gnuplot, and this problem is not evident there. So it appears to me the problem is in the DotPlotly script somewhere.

I am running Mummer 4.0, with the following commands:
Nucmer 1.c5.fasta 2.c4.fasta
show-coords -c out.delta > out.coords

It seems this problem results anytime I set DotPlotly -m to 5kb or higher, but when -m is 1000 or less, this does not occur.

Any fix or advice would be helpful, I would really like to use Dotplotly, but am concerned this problem may extend to other contigs in a larger analysis.

Thank you,
Jon

Mauve vs Nucmer.pdf