Multiple Alignment
I work with several people on problems related to
estimating multiple alignments.
- Jim Leebens-Mack, Claude de Pamphilis,
Li-San Wang and I have been looking at the question of
whether improving the alignment results in improved phylogenetic
estimations, using standard phylogenetic estimation techniques
(including maximum parsimony, maximum likelihood, and distance-based
methods) and standard MSA methods (including MAFFT, ProbCons,
Muscle, ClustalW, and other methods).
- Randy Linder, Kevin Liu, David Zhao and I have been
developing new heuristics for the problem of finding minimum
length trees, under affine gap penalties. Our work was
presented as a
poster at PSB 2007.
This heuristic requires the user to specify
a gap penalty function.
-
Our next objective will be to develop a statistical estimation technique for
simultaneous estimation of trees and alignments, which
can scale to hundreds of sequences.