Question re -fconst option

[Megan Earls]

Hi,

I am running IQTree on the snippy-core output (core.aln) of ~200 staph aureus genomes. As I am using an alignment with the SNP sites only, I understand that I must use the -fconst option in order for my branches to be scaled correctly (is this correct?). 

My question is, how do I know what integer numbers to put after "-fconst" in my command? I understand that "-fconst 10, 20, 15, 40" relates to "-fconst A, C, G, T" but how do I know what exact numbers to use?

My SNP site alignment is ~8000 nt and the staph aureus genome is ~2.8 MB, with a GC content of ~32%

Any help would be really appreciated.

Thanks,

Megan

[Bui Quang Minh]

Hi Megan,

On Fri, 6 Mar 2020 at 11:45 pm, Megan Earls <ear...@tcd.ie> wrote:

Hi,

I am running IQTree on the snippy-core output (core.aln) of ~200 staph aureus genomes. As I am using an alignment with the SNP sites only, I understand that I must use the -fconst option in order for my branches to be scaled correctly (is this correct?). 

Yes that’s correct, but Iq-tree can still do ascertainment bias correction automatically to account for the absence of invariant sites.

My question is, how do I know what integer numbers to put after "-fconst" in my command? I understand that "-fconst 10, 20, 15, 40" relates to "-fconst A, C, G, T" but how do I know what exact numbers to use?

My SNP site alignment is ~8000 nt and the staph aureus genome is ~2.8 MB, with a GC content of ~32%

Any help would be really appreciated.

That’s I don’t know, perhaps this snippy program can tell you with some option? In case that’s not possible, the ASC model mentioned above can be used

Minh

Thanks,

Megan

--
You received this message because you are subscribed to the Google Groups "IQ-TREE" group.
To unsubscribe from this group and stop receiving emails from it, send an email to iqtree+un...@googlegroups.com.
To view this discussion on the web visit https://groups.google.com/d/msgid/iqtree/802904d7-de67-49d4-95c6-da5485d1f507%40googlegroups.com.

[Megan Earls]

Hi Minh,

Thank you for your response. I will check snippy for that detail. Re the +ASC option, I ran the following command:

iqtree -s core.aln -m GTR+ASC

And the tree turned out pretty much as expected. But how did the branch scaling get corrected when the I never told IQTree the number of invariant site or the size of the genome?

Thanks,

Megan

[Minh Bui]

Hi Megan,

This “mystery” was explained in a great paper by Paul Lewis:

https://doi.org/10.1080/106351501753462876

IQ-TREE implemented exactly this correction.

Cheers,

Minh

-- 
You received this message because you are subscribed to the Google Groups "IQ-TREE" group.
To unsubscribe from this group and stop receiving emails from it, send an email to iqtree+un...@googlegroups.com.

To view this discussion on the web visit https://groups.google.com/d/msgid/iqtree/284bb805-4634-4162-8f37-71e163a497fe%40googlegroups.com.

[Megan Earls]

Thank you!

On Friday, March 6, 2020 at 12:45:39 PM UTC, Megan Earls wrote: