Detecting synteny among multiple genomes?

[aminakur]

I want to use progressiveCactus to align multiple genomes and produce a Hal file. Is it possible to extract then blocks of synteny shared between several genomes, not just two? Seems like halSynteny only allows comparing two genomes.

[PEHGP]

+1

[glennhickey]

I think the only way to do this is through MAF. hal2maf won't be of much direct help as the blocks are too fragmented, but you can use cactus-hal2maf and/or taffy (included in cactus) to merge up MAF blocks into larger, syntenic blocks.

[brantfaircloth]

Hi Glenn,

Just a quick question since it seems on-topic here... is cactus-hal2maf (or taffy, but it basically looks like cactus-hal2maf is using taffy norm) the preferred method of extracting these regions compared to an older tool like halper (which is a little bit of a bugaboo to run, uses an older HAL API, etc.)?

I've found the recent version of cactus-hal2maf (v2.8.2) to work nicely with an input BED file of regions to extract to MAF (thank you!), but I haven't gone down the road of comparing different methods that do approximately the same thing. So, I realize this is a bit of a lazy question, but I thought you might have a quick response.

[glennhickey]

cactus-hal2maf is definitely the best way to get MAF. For pairwise alignments you can try cactus-hal2chains with or without --useHalSynteny that will do some additional synteny filtering that the MAF export doesn't do (yet, at least).

[brantfaircloth]

Cool, thanks 👍 - I'll give that a test.