tag:blogger.com,1999:blog-8386055846297828307.post7738889593764473714..comments2018-02-14T17:10:19.019+01:00Comments on The Genealogical World of Phylogenetic Networks: Clades, cladograms, cladistics, and why networks are inevitableDavid Morrisonnoreply@blogger.comBlogger3125tag:blogger.com,1999:blog-8386055846297828307.post-32200350002172584352018-01-28T08:48:03.024+01:002018-01-28T08:48:03.024+01:00I think I will give a try with these "median ...I think I will give a try with these "median networks". Thank you.dauberthttps://www.blogger.com/profile/17724042372602455356noreply@blogger.comtag:blogger.com,1999:blog-8386055846297828307.post-44447451325201083902018-01-17T13:42:04.503+01:002018-01-17T13:42:04.503+01:00Just saw the comment.
The commagram is a nice gr...Just saw the comment. <br /><br />The commagram is a nice graph, it reminds me of phylogenetic trees before Hennig (there are quite a bunch of them in the classic literature).<br /><br />In principle, you go in the same direction: phylogenetic trees that include non-coeval taxon should allow placing OTUs at nodes or along branches, but the technical problem of encoding them remains. The NEWICK format is essentially cladistic.<br />(I managed to trick Wikipedia's cladogram format to include paraphyla by making use of the branch label option: <a href="https://en.wikipedia.org/wiki/Osmundaceae" rel="nofollow">https://en.wikipedia.org/wiki/Osmundaceae</a>)<br /><br />Regarding the remaining aspects: the median network already optimises ancestral-descendant relationships. So I would try to find ways to handle homoplasy (via character filtering/weighting) to make it possible to run a median-joining/reduced-median network on the morphological data, and not optimising a cladogram that better fits the expected "caulogram". You may give it a try with your data, and your Bayesian weights. You could categorise/rank them and apply them when running the median network analysis. <a href="http://www.fluxus-technology.com/sharenet.htm" rel="nofollow">NETWORK's</a>, the free median network software, matrix edit option allow assigning a weight to each column.<br /><br />However, one problem with median networks may be scalability (working with large taxon sets), we have fast parsimony tree-inference implementation but not for fast median-network-inference (maybe impossible?!) But this is only a theoretical problem. <br /><br />In praxis, for histories including fossils, time-slice-limited stacked median networks could be an easy work-around. But maybe I'm too sceptical regarding all-inclusive reconstructions covering millions and millions of years (and evolution). So there may be a market for parsimony tree methods optimising internal OTUs.Das Grimmhttps://www.blogger.com/profile/13213125622809796109noreply@blogger.comtag:blogger.com,1999:blog-8386055846297828307.post-8455735681209725772017-12-16T15:55:39.381+01:002017-12-16T15:55:39.381+01:00My own work could be of some interest because it i...My own work could be of some interest because it is somewhat related to the incompleteness and distortion introduced by cladograms : https://www.researchgate.net/publication/317429868_A_simple_parsimony-based_approach_to_assess_ancestor-descendant_relationshipsdauberthttps://www.blogger.com/profile/17724042372602455356noreply@blogger.com