RAxML NG : custom amino acid substitution model

295 views
Skip to first unread message

Thomas Ed. Bigot

unread,
Feb 19, 2019, 11:57:26 AM2/19/19
to raxml
Hi,

Is it possible with raxml-ng to use a custom amino-acid model?
I’d like to specify a transition matrix (in paml format for instance), like one can do with paml (http://www.iqtree.org/doc/Substitution-Models#protein-models § User-defined empirical protein models
)

Thanks in advance,
Thomas

Alexey Kozlov

unread,
Feb 19, 2019, 12:09:12 PM2/19/19
to ra...@googlegroups.com
Hi Thomas,

yes it is possible, although the format is slightly different:

https://github.com/amkozlov/raxml-ng/wiki/Input-data#single-model

e.g. something like "--model PROTGTR{rates.txt}+FU{freqs.txt}" - please
note that rates are in different order (upper triangle)!

PAML format support is on my TODO list however, and it will become
available in the next release I guess.

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.

Thomas Ed. Bigot

unread,
Feb 20, 2019, 5:10:53 AM2/20/19
to raxml
Thanks a lot for your reply.

For protein model, what order is expected? Is it, as for PAML, alphabetical order of three letter aminoacid codes (“Ala Arg Asn Asp Cys Gln Glu Gly His Ile Leu Lys Met Phe Pro Ser Thr Trp Tyr Val”) or the alphabetical order of IUPAC one letter code ?

Thank in advance,
Thomas

Alexey Kozlov

unread,
Feb 20, 2019, 5:35:01 AM2/20/19
to ra...@googlegroups.com
Hi Thomas,

aminoacids are in the following order:

https://github.com/amkozlov/raxml-ng/wiki/Input-data#defaults

Best,
Alexey

Thomas Ed. Bigot

unread,
Feb 21, 2019, 7:02:36 AM2/21/19
to raxml
So I can notice it matches the alphabetical order of three letter codes, like in Paml.
Thanks !

Cheers,
Thomas

Le mardi 19 février 2019 17:57:26 UTC+1, Thomas Ed. Bigot a écrit :

Alexey Kozlov

unread,
Feb 21, 2019, 7:51:09 AM2/21/19
to ra...@googlegroups.com
yes, I guess that's how this particular order came about...

Alexey
Reply all
Reply to author
Forward
0 new messages