The limits of liftOver

[David Garfield]

Gruß,

In my current project, I’m working with individuals of the same species plus a few closely related species. 

I know that liftOver is generally recommended for transitions between assemblies, while pslMap is certainly recommended for between more distant species.

But at what point would you suggest switching from one tool to the other? When do two genomes stop being “assembly”-like, and start being species-like?

Cheers,

David

-------------------------------------------------------------------------------------

David Garfield, PhD

Furlong Group

European Molecular Biology Laboratory (EMBL)

Telephone    +49 6221 387 8426

Fax                 +49 6221 387 166

E-mail david.g...@embl.de

Snail Meyerhofstraße 1

D-69012 Heidelberg

Germany

[Luvina Guruvadoo]

Hello David,

Thanks for your question. Here are a few comments from our engineers:

liftOver maps the start and end positions independently, so if there is a rearrangement between the two sequences, the number of bases in the region can change dramatically. pslMap forces a block structure on the alignment so the number of bases in the alignment is the same in both sequences (though there may be very large "introns", and inversions and translocations will cause blocks to be dropped because they aren't in the same order and orientation any more).

I prefer pslMap because it's a base to base mapping. liftOver is useful if you just want to know where the same region is in a
different assembly of the same individual which presumably isn't rearranged with itself but may have inserted bases where there was a gap before (though obviously translocations do occur within cells of individuals...hello cancer). Both procedures use the same chain or psl file, which can be made by lastz or blat or etc.The author of pslMap also had this to say:

Both methods are dependent on the alignments used and how they are filtered. The README in pslMap describes how to do the syntenic filtering used for cross-species transmap. It requires some experimentation to see if the alignments produce the desired results.

The bit algorithmic difference is that liftover lifts block boundaries which pslMap is per-base. So liftover produces something more resembling an annotation while pslMap shows the evolutionary changes. Also, pslMap requires converting to psl; there are tools to do this.

I hope this helps. If you have any further questions, please reply to gen...@soe.ucsc.edu. All messages sent to that address are archived on a publicly-accessible forum. If your question includes sensitive data, you may send it instead to genom...@soe.ucsc.edu.

- - -
Luvina Guruvadoo
UCSC Genome Bioinformatics Group

--