Discuss how to set arguments for equal/ML (DNA) model/empirical (AA). ModelTest is intended to provide flexibility for including only one set of frequencies, or both.
ProtTest and old ModelTest versions include the equal(DNA)/model(AA) frequencies by default. ML(DNA)/empirical(AA) are included when the argument -F is specified. However, the new version of modeltest uses an argument such as -F [e|m]. For example:
-F e
-F m
-F em
'e' can stand for equal/empirical and 'm' for ML/model. The problem is that one could mix up both, because 'e' increases the number of parameters for protein models, while 'm' does the same for DNA models. If only one single data type is used, this might not be a big deal. However, if the input is a partitioned data set including both DNA and protein data, the user won't be able to choose only those models with or without free parameters on the frequencies.
Other proposal is to use, for example, 'f' and 'e' standing for 'fixed' and 'estimated', instead of 'e' and 'm'. Therefore, 'f' will be used for equal(DNA)/model(AA); and 'e' for ML(DNA)/empirical(AA):
Discuss how to set arguments for equal/ML (DNA) model/empirical (AA). ModelTest is intended to provide flexibility for including only one set of frequencies, or both.
ProtTest and old ModelTest versions include the equal(DNA)/model(AA) frequencies by default. ML(DNA)/empirical(AA) are included when the argument -F is specified. However, the new version of modeltest uses an argument such as -F [e|m]. For example:
'e' can stand for equal/empirical and 'm' for ML/model. The problem is that one could mix up both, because 'e' increases the number of parameters for protein models, while 'm' does the same for DNA models. If only one single data type is used, this might not be a big deal. However, if the input is a partitioned data set including both DNA and protein data, the user won't be able to choose only those models with or without free parameters on the frequencies.
Other proposal is to use, for example, 'f' and 'e' standing for 'fixed' and 'estimated', instead of 'e' and 'm'. Therefore, 'f' will be used for equal(DNA)/model(AA); and 'e' for ML(DNA)/empirical(AA):