cactus-hal2maf: are regions on different scaffolds considered dupes?

[gwct]

Hi,
I've recently observed that when running cactus-hal2maf with --outType single I still occasionally get MAF blocks with >1 sequence for a given species. I haven't done a thorough check yet, but it seems that these 'duplicates' occur on different scaffolds from one another within that species' assembly. Does cactus-hal2maf not consider sequences on different scaffolds within an assembly to be duplicates ever? Or (more likely) am I misunderstanding something about how cactus-hal2maf is working here.

I can provide examples if needed, but I thought I would ask in case the answer was something simple that I wasn't getting.

Thanks!

[glennhickey]

--outType single should send things through mafDuplicateFilter. If that's not working properly it's likely due to something in the sequence names confusing it. I think cactus is supposed to try to temporarily rename things to avoid confusion, so perhaps there's a problem in this process... Anyway, feel free to send an example...