Replies: 2 comments
-
|
@thomaskf Does partition models work with mixture models? (i.e., each partition might follow a mixture model?) I need to check this first. Thanks. |
Beta Was this translation helpful? Give feedback.
-
|
I've got some results from MF+MERGE which successfully found the best partitioning scheme for a set of 28 alignments using mixture models. -p suggested combining them into 13 partitions, each modelled under C60 with G and Q.pfam, Q.bac or LG. Command: iqtree2 -p Partitions.nex -m MF+MERGE -mrate E,I,G,I+G,R -madd LG+G+F+C60,LG+G+C60,Q.pfam+G+F+C60,Q.pfam+G+C60,Q.bac+G+F+C60,Q.bac+G+C60 -pre Mixture_p_merged -ntmax 15 -nt AUTO -safe -rcluster 5 Log file: Mixture_p_merged.log Key part of the best_scheme.nex: charpartition mymodels = |
Beta Was this translation helpful? Give feedback.
-
Hi IQ-TREE Users,
Do any of you have any more insight on what this warning means and how to avoid it?
"WARNING: Too high (saturated) partition rates for proportional partition model!"
It came-up when I ran a partitioned analyses using the -p option to allow each partition to evolve at it's own rate. The partitions are admittedly quite large - there are 13 altogether, each containing between 57 and 1,967 alignment columns and up to 1,025 sequences. They're also modelled using complex mixture models (a combination of the mixture model C60 with Q.pfam, Q.bac or LG and G), so I'm aware it's a complex analyses with many parameters.
My question is can I trust the resulting phylogeny when partition rates aren't being estimated properly? I'm assuming the answer is no, but I wonder if iq-tree defers to a simpler model like -q when partition rates are too high?
Log file here so you can see where the warning appears: PartitionWarning.log
Beta Was this translation helpful? Give feedback.
All reactions