Skip to content

BASEMLG

sabifo4 edited this page Nov 1, 2024 · 3 revisions

BASEMLG

This PAML program uses the same control file baseml.ctl as the PAML program BASEML. Tree-search or the assumption of a molecular clock are not allowed, and so you need to specify runmode = 0 and clock = 0 in the control file. Substitution models available for BASEMLG are JC69, F81, K80, F84, and HKY85, and a continuous gamma is always assumed for rates at sites. The variables ncatG, fix_rho, rho, and nhomo have no effect. The S.E.'s of parameter estimates are always printed out because they are calculated during the iteration, and so getSE has no effect.

Because of the intensive computation required by BASEMLG, the discrete-gamma model implemented in BASEML is recommended for data analysis. If you choose to use BASEMLG, you should run BASEML first, and then run BASEMLG. This allows BASEML to collect initial values into a file named in.basemlg, which you can then use as input sequence file for BASEMLG. Note that BASEMLG implements only a subset of models compared to BASEML (i.e., JC69, F81, K80, F84, and HKY85), and so you should run your first analysis in BASEML under one of such models if you then want to use BASEMLG.

Clone this wiki locally