incorporation of known Q matrix into global model for AA data

[Jenna Mccullough]

Hi, 

I'm trying to use the Q.bird substitution matrix for a RAxML-ng analysis and i am getting errors saying that it is not a correct model. 

This is the command line, which i am running on my University's HPC cluster.

raxml-ng --threads 64 --all --msa todi101_busco8068_concat.phylip --bs-trees autoMRE{100} --model LG+Q.bird

Program version RAxML-NG v. 1.0.1 

The error is: ERROR: Wrong model specification: +Q.bird 

What am i doing wrong for this? The manual (https://github.com/amkozlov/raxml-ng/wiki/Input-data#evolutionary-model) says to add LG with the additional matrix command after "+". 

Thank you 

Jenna 

[Oleksiy Kozlov]

Hi Jenna,

Q.bird matrix is supported since raxml-ng version 1.2.0:

https://github.com/amkozlov/raxml-ng/releases/tag/1.2.0

With earlier versions, you can provide Q.bird matrix in the PAML format and use "--model
PROTGTR{qbird.txt}":

https://giphy.pasteur.fr/empirical-models-of-amino-acid-substitution/Q.bird.dat

Hope this helps,
Oleksiy

> --
> 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/1d5a50f4-c70a-450e-afed-37152004d458n%40googlegroups.com
> <https://groups.google.com/d/msgid/raxml/1d5a50f4-c70a-450e-afed-37152004d458n%40googlegroups.com?utm_medium=email&utm_source=footer>.

[Jenna Mccullough]

Hi, Oleksiy,

Thank you for your help. As you suggested, I first tried running it with the earlier version of raxml-ng (specifying "--model PROTGTR{qbird.txt}”) and I still received the error:

"ERROR: Invalid substitution rate specification: {qbird.txt}” 

So I updated the program to 1.2.0, changed the model to be LG+Q.bird (as suggested in the manual) and still received the error:  

ERROR: Wrong model specification: +Q.bird

Just for my own curiosity, I then tried with the model specified as "PROTGTR+Q.bird” and that still had the error: 

ERROR: Wrong model specification: +Q.bird

So then I just tried it with “--model  Q.bird” and it worked. 

I suggest that the manual (https://github.com/amkozlov/raxml-ng/wiki/Input-data#evolutionary-model) be slightly edited to reflect this, as I definitely misinterpreted the phrase: "Model specification always starts with a substitution matrix name, e.g., GTR for DNA data or LG for protein data. Several optional modifiers can be added, separated by + and in arbitrary order.” To me, that means that any protein model must start with “LG” and anything else must be added with a “+”. 

Regardless, thank you for your help Oleksiy

Jenna 

You received this message because you are subscribed to a topic in the Google Groups "raxml" group.
To unsubscribe from this topic, visit https://groups.google.com/d/topic/raxml/T5vMPObMos8/unsubscribe.
To unsubscribe from this group and all its topics, send an email to raxml+un...@googlegroups.com.
To view this discussion on the web visit https://groups.google.com/d/msgid/raxml/e613b066-8501-986c-9925-7f8a28dc35b9%40gmail.com.