misidentification correction

[Krystyna]

Hi,

I am planning to run dadi and having read the manual one thing is not
exactly clear to me:
Are expected number of substitutions, trinucleotide mutation rate
matrix and stationary trinucleotide distributions needed to be
calculated before applying correction for multiple mutations in dadi
or dadi does it when running Spectrum.from_data_dict_corrected?

Thanks for help in advance!

Krystyna

[Gutenkunst, Ryan N - (rgutenk)]

Hello Krystyna,

The expected number of substitutions, trinucleotide mutation rate matrix,
and stationary trinucleotide distributions do need to be calculate outside
of dadi. (Although we distribute a mammalian trinucleotide substitution
matrix.) The number of substitutions and stationary trinucleotide
distribution should be easy to calculate by comparing your outgroup
sequence with the reference sequence of the organism you're modeling.

Best,
Ryan

--
Ryan Gutenkunst
Assistant Professor
Molecular and Cellular Biology
University of Arizona
phone: (520) 626-0569
http://gutengroup.mcb.arizona.edu

[Aryn Wilder]

Hi Ryan,

Thanks for the great program and forum.

I am trying to correct my AFS for ancestral misidentification. Where can I find the mammalian trinucleotide substitution matrix that you distribute? In your post to Krystyna, are you referring to the Q.HwangGreen.human.dat file in the dadi package? If so, is that matrix human specific or do you think it could apply to mammals in general?

Thanks for your help,
Aryn

[Gutenkunst, Ryan N - (rgutenk)]

Hello Aryn,

In the dadi source distribution, the human matrix we distribute is in the directory examples/fs_from_data .

Hwang and Green did infer different matrices for different mammalian clades, so using the correct one for your species would be best. I only have the human matrix, so for the other clades, you’ll have to ask them.

Best,

Ryan

--
You received this message because you are subscribed to the Google Groups "dadi-user" group.
To unsubscribe from this group and stop receiving emails from it, send an email to dadi-user+...@googlegroups.com.
To post to this group, send email to dadi...@googlegroups.com.
Visit this group at http://groups.google.com/group/dadi-user.
For more options, visit https://groups.google.com/groups/opt_out.