molec evo quiz 7

upgma uses

pairwise distances - u

neighbor joining uses

pairwise distances - n

maximum parismony uses

characters directly, -m

maximum liklihood uses

characters directly -ml

bayesian inference uses

characters dircetly -b

upgma chooses tree by

clustering, joining the closest pair, repeat, no alternate trees are scored

neighbor joining chooses tree by

clustering, join the lowest q. a fast apprx to to minimum evolution

maximum parismony chooses tree by

criterion + heurisitc serarch, fewest charcater changes

maximum liklihood chooses tree by

criterion + heuristic search, highest P (data | tree, model)

bayesian inference chooses trees by

criterio + heuritc search; from posterior

upgma model of evolution

only possible isnide the distance correction

neighbor joining model of evolution

only possible inside the distance correction

maximum parisomony model of evolution

none

maximum liklihood model of evolution

explicit

bayesian inference model of evolution

explicit, plus priors

upgma branch lenghts

forced euqal at every join so tree is ultrametric

neighbor joining branch lengths

estimated, returns an unrooted tree

maximum parisomony branch lenghts

based on inferred character changes

maximum liklihood branch lengths

estimated frmo data under a model

bayesian inference branch lengths

estimated with uncertaininty

upgma and neighbor joining speed

superfast

maximum parisomny speed

fast

maximum liklihood speed

medum to slowb

bayesian inference speed

slow

main weakness of upgma

molecular clock, groups by similarlity, not ancestry, discards site via matrix

main weakness of neighborjoining

discards ste via matrix

main weakness of maximum parisomony

long branch attraction, doesnt use model of evolution

main weakness of maximum liklihood

only as gooda s the modelma

main weakness of bayesian inference

sensitive to priors, convergence has to be checked