These are chat archives for dereneaton/ipyrad

Aug 2017
Glib Mazepa
Aug 03 2017 14:15
Hi folks, I have quite annoying issue with phylogenetic reconstruction with ipyrad: the dataset (alignment of 300k bp, clustering threshold - 0.85, 100 indivs and minimum coverage per locus - 30 indivs) was constructed by denovo+reference approach. The taxa set spans very deep divergencies (40 Mya or more) as well as very shallow splits. I checked the dataset with ADMIXTURE and removed putative hybrids and then ran BEAST and MrBayes - I get two alternative topologies that are well supported and drastically differ in the position of the focal taxa group: these topologies are constructed with BEAST with more and less complete supermatrixes, and MrBayes produces them on the same input in different chains)... Any suggestions? I would try partitioning, but unfortunately it is not implemented unlike in pyrad... Should i try to raise the clusterig threshold perhaps?