--asc-corr=stamatakis how to provide number of invariant sites

[Sean]

Hello,

I am working with RAD data and have a phylip file containing only variable sites. How do I set the number of invariant sites using the correction --asc-corr=stamataki ? The manual says to use the partition file. So would I calculate the number of total sites (T) and just create a partition file with DNA, p1:1-T? (and the invariant sites will be inferred?)

Also, what is the best way to calculate the number of invariable sites from a vcf file?

Thanks,
Sean

[Alexey Kozlov]

Hi Sean,

> I am working with RAD data and have a phylip file containing only variable sites. How do I set the number of invariant
> sites using the correction --asc-corr=stamataki ? The manual says to use the partition file. So would I calculate the

> number of total sites (T) and just create a partition file with /DNA, p1:1-T/? (and the invariant sites will be inferred?)

T must be the number of (variable) sites in your PHYLIP file. Please see p. 44 of RAxML manual for the explanation how
to specify the number of invariant sites:

https://sco.h-its.org/exelixis/php/countManualNew.php

> Also, what is the best way to calculate the number of invariable sites from a vcf file?

VCF file usually contain variable sites only (SNPs), so I don't think there is any possibility to *calculate* the number
of invariant sites. However, this information could be theoretically specified in one of the header fields (but not sure
about this).

Best,
Alexey

[Mariana López]

Hi!!

I need some help, actually I need to check if I am interpreting correctly the option ASC_STAM:

I work with a genome size specie of 4.4 millions bp (or positions), I made whole genome sequencing, but after the SNP identification pipeline, I have an alignment.fasta file with ONLY variable sites (18000 variable positions)

If I am right, when usign ASC_FELS the number of invariant sites will be (4382000=4400000-18000), is that ok??

If I used instead ASC_STAM, I must define the number of those 4382000 positions that are A/C/G/T?? Is that ok?? I know the proportion of each nucleotide in the genome, but I must define the total variable As and so on...

I am asking this, because I discussed the parameters used by a collegue, and he defined the invariant sites as the total position of the whole genome that are A/C/G/T, not considering the variable sites.. I mean he reported all the genome as invariant sites plus the variable sites present in the fasta file. Because of that I need to confirm if I am rigth or wrong!!

tranks in advance!

sincerely

Mariana

[Alexey Kozlov]

Hi Mariana,

- for the Felsenstein correction, you put the *total* number of *invariable* sites, i.e. ASC_FELS{4382000}

- for the Stamatakis correction, you put the number of *invariable* sites with As, Cs, Gs, and Ts, i.e.
ASC_STAM{nA/nC/nG/nT}, where nA + nC + nG + nT = 4382000

Since you have whole-genome sequences, all those numbers can be easily computed by simple character counting.

Best,
Alexey

> I work with a genome size specie of 4.4 millions bp (or positions), I made whole genome sequencing, but after the SNP
> identification pipeline, I have an alignment.fasta file with ONLY variable sites (18000 variable positions)
> If I am right, when usign ASC_FELS the number of invariant sites will be (4382000=4400000-18000), is that ok??
> If I used instead ASC_STAM, I must define the number of those 4382000 positions that are A/C/G/T?? Is that ok?? I know
> the proportion of each nucleotide in the genome, but I must define the total variable As and so on...
>
> I am asking this, because I discussed the parameters used by a collegue, and he defined the invariant sites as the total
> position of the whole genome that are A/C/G/T, not considering the variable sites.. I mean he reported all the genome as
> invariant sites plus the variable sites present in the fasta file. Because of that I need to confirm if I am rigth or
> wrong!!
>
> tranks in advance!
> sincerely
> Mariana
>
>
> El domingo, 15 de abril de 2018, 20:22:13 (UTC+2), Alexey Kozlov escribió:
>
> Hi Sean,
>
> > I am working with RAD data and have a phylip file containing only variable sites. How do I set the number of
> invariant
> > sites using the correction --asc-corr=stamataki ? The manual says to use the partition file. So would I calculate
> the
> > number of total sites (T) and just create a partition file with /DNA, p1:1-T/? (and the invariant sites will be
> inferred?)
>
> T must be the number of (variable) sites in your PHYLIP file. Please see p. 44 of RAxML manual for the explanation how
> to specify the number of invariant sites:
>

> https://sco.h-its.org/exelixis/php/countManualNew.php <https://sco.h-its.org/exelixis/php/countManualNew.php>

>
> > Also, what is the best way to calculate the number of invariable sites from a vcf file?
>
> VCF file usually contain variable sites only (SNPs), so I don't think there is any possibility to *calculate* the
> number
> of invariant sites. However, this information could be theoretically specified in one of the header fields (but not
> sure
> about this).
>
> Best,
> Alexey
>

> --
> You received this message because you are subscribed to the Google Groups "raxml" group.
> To unsubscribe from this group and stop receiving emails from it, send an email to raxml+un...@googlegroups.com
> <mailto:raxml+un...@googlegroups.com>.
> For more options, visit https://groups.google.com/d/optout.

[Mariana López]

Thanks in advance!!

Mari

To unsubscribe from this group and stop receiving emails from it, send an email to raxml+unsubscribe@googlegroups.com <mailto:raxml+unsubscribe@googlegroups.com>.

For more options, visit https://groups.google.com/d/optout.

--
You received this message because you are subscribed to the Google Groups "raxml" group.

To unsubscribe from this group and stop receiving emails from it, send an email to raxml+unsubscribe@googlegroups.com.

For more options, visit https://groups.google.com/d/optout.

--

Los grandes espíritus siempre han encontrado una violenta oposición de parte de mentes mediocres. A. Einstein

[Alexandros Stamatakis]

Are you using RAxML-NG or standard RAxML?

Would be helpful if you could paste in the command line and the exact
error you are getting.

Alexis

On 11.05.19 17:04, Hollie Topliffe wrote:
> Hi,
>
> Sorry to jump on this thread - but I'm trying to do the same thing with
> probably the same type of data (SNP data from TB genomes!) but when I
> specify the ASC_STAM parameter it tells me this is no longer possible.
> How do you go about this with the "-m ASC_GTRGAMMA --asc_corr=stamatakis
> -q txtfile" setup? Do the number of invariable sites go in to the txt
> file specified by -q? How is this formatted in the file?
>
> Thanks,
> Hollie

> send an email to ra...@googlegroups.com <javascript:>
> > <mailto:ra...@googlegroups.com <javascript:>>.

> > For more options, visit https://groups.google.com/d/optout

> <https://groups.google.com/d/optout>.

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

> To view this discussion on the web visit
> https://groups.google.com/d/msgid/raxml/d40884f5-eb42-49cc-acfe-330934079b9c%40googlegroups.com
> <https://groups.google.com/d/msgid/raxml/d40884f5-eb42-49cc-acfe-330934079b9c%40googlegroups.com?utm_medium=email&utm_source=footer>.

> For more options, visit https://groups.google.com/d/optout.

--
Alexandros (Alexis) Stamatakis

Research Group Leader, Heidelberg Institute for Theoretical Studies
Full Professor, Dept. of Informatics, Karlsruhe Institute of Technology

www.exelixis-lab.org