Site Model: Query Regarding Codon Bias in Organisms

[Upasana Maity]

Hello,

I am running CODEML site model to identify positively selected genes. Given that our alignment file includes multiple organisms, each with their own codon usage patterns, I am concerned about how CODEML addresses codon bias in this context. How does CODEML account for the differences in codon usage among the various organisms in the alignment file during its analysis?

Thank you in advance.

Sincerely,

Upasana

[Sandra AC]

Hi there,

By searching for "codon usage bias" in the search bar available in this google group, you can find this old post in the PAML discussion group that may help answer your query.

In addition, you may want to check option `CodonFreq` in the PAML documentation to decide which setting best fits your data. This option can enable different codon frequencies for different codons. If the species or genes are so different that the different sequences in the alignment have different frequencies (e.g., some sequences are AT-rich while other sequences are GC-rich), then one should not conduct codon-based analyses -- this may happen when the species are distantly related, and the sequences are very divergent.

Hope this helps!
S.