Conservation Scoring Across Motif Sites

[alexjironkin]

Hi,

I am trying to figure out how to compute conservation of sequence in and around motif across multiple sites, but within the same species, the same as the conservation plot here:

http://www.nature.com/nature/journal/v489/n7414/images/nature11212-f1.2.jpg

(C, middle)

They used phyloP. However, that requires model of substitution, e.g. neutral.mod. That can be done via phyloFit, but I am only interested in 1 species, in this case Arabidopsis and I am not sure what to use to compute model! I don't think using coding sequence of homologs is right, because I am looking into promoter sequences. So what I was trying to do is to use all of my sites for a motif(~1000) each about 50 bp long to estimate this model.

As soon as I run phyloFit with 4 FASTA sequences it complains:
"ERROR: --tree required"

So I am not sure how to generate model using phyloFit and multiple sequences from the same species.

Then, let's say I have model, I also have aligned my 50bp sites using some multiple aligner and run:

phylop model.mod in.fasta
Returns:
ERROR: no match for leaves of tree in alignment.


I don't understand how it relates leaves in the tree that is generated by phyloFit and the inputs to phyloP. Any help would be greatly appreciated.


Alex

[balaji]

Can you list the sequence of commands you tried.
What commands did you give for phyloFit
and also you say
"phylop model.mod in.fasta"
how did you generate the model.mod file that you are using as an input to phylop

--
phyloFit in.fasta (generates model.mod)
the model.mod can then be used in phylop model.mod in.fasta

[alexjironkin]

Hi,

I don't have .mod file yet because I don't want model across species. I want model within arabidopsis, if you see what I mean? I am only going to be comparing sites form arabidopsis genes so I need substitution/deletion/etc rates within genome. As the first step I am not sure how to generate that!

Any ideas?

Alex