Discrepancy between liftOver results and Multiz Alignments track?

[Menglan Xiang]

Hi there,

We are interested in genomic regions conserved between mouse and human. I have some candidate regions in the mouse genome that when manually examined in the genome browser appear to be highly conserved (according to the Multiz Alignments track). I’m interested in finding where the corresponding regions are in the human region, so I used the liftOver tool with the mm10 to hg19 chain file.

These are the mm10 regions I used for testing:

chr7:30636217-30636221
chr7:30636229-30636233
chr5:119837363-119837367


Which are converted to the following regions in hg19:

chr19 35642040 35642044 mouse_NR2F2_chr7:30636217-30636221
chr19 35642028 35642032 mouse_NR2F2_chr7:30636229-30636233
chr12 114403366 114403370 mouse_NR2F2_chr5:119837363-119837367


However, these human regions not only not have the same sequence as the syntenic mouse regions, but the aligned mouse sequence are either not matching the human sequence or nonexistent. So I’m wondering how Multiz Alignments track is generated and does it work differently than liftOver?

Thank you,
Menglan Xiang

[Gerardo Perez]

Hello, Menglan.

Thank you for your interest in the Genome Browser and your question about your liftOver results.

We are not able to replicate your converted regions using liftOver in hg19. We did match your converted regions in hg38:

chr19  35642040  35642044  chr7:30636218-30636221    
chr19  35642028  35642032  chr7:30636230-30636233    
chr12  114403366  114403370  chr5:119837364-119837367

When converting your mm10 regions using liftOver to hg19, we get the following:

chr19 36132942    36132947    mouse_NR2F2_chr7:30636217-30636221
chr19 36132930    36132935    mouse_NR2F2_chr7:30636229-30636233
chr12 114841171   114841176   mouse_NR2F2_chr5:119837363-119837367

We do see that your regions land at inversion locations. When there is an inversion, the liftOver chain has a block on the - strand. For example, using the region chr7:30,636,217-30,636,221 on mm10, go to the following session then click the purple chain "chr19 - 30087k", you will see “Strand: -”:
http://genome.ucsc.edu/s/gperez2/mm10_chr7%3A30%2C636%2C217-30%2C636%2C221

Since the regions land at inversion locations, you can reverse the liftOver region sequence to see the same sequence as the Multiz Alignments track. You can click the “reverse” button to reverse the sequence. The following session has the hg19 chr19:36,132,943-36,132,947 sequence region reversed and matches the sequence in the previous session:
http://genome.ucsc.edu/s/gperez2/hg19_chr19%3A36%2C132%2C943-36%2C132%2C947

I hope this is helpful. 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 Google Groups forum. If your question includes sensitive data, you may send it instead to genom...@soe.ucsc.edu.

Gerardo Perez
UCSC Genomics Institute

--

---
You received this message because you are subscribed to the Google Groups "UCSC Genome Browser Public Support" group.
To unsubscribe from this group and stop receiving emails from it, send an email to genome+un...@soe.ucsc.edu.
To view this discussion on the web visit https://groups.google.com/a/soe.ucsc.edu/d/msgid/genome/E517FD70-A206-4BF2-9D4F-0D3482FEFFBA%40stanford.edu.