Summary: Maximal Accurate Forests from Distance
Matrices
C. Daskalakis, C. Hill, A. Ja#e,
R. Mihaescu, E. Mossel, S. Rao
University of California, Berkeley
Research supported by CIPRES (NSF ITR grant # NSF EF 03­31494)
Abstract. We present a fast converging method for distance­based phy­
logenetic inference, which is novel in two respects. First, it is the only
method (to our knowledge) to guarantee accuracy when knowledge about
the model tree, i.e bounds on the edge lengths, is not assumed. Second,
our algorithm guarantees that, with high probability, no false assertions
are made. The algorithm produces a maximal forest of the model tree,
in time e
O n 3


in the typical case. Empirical testing has been promising,
comparing favorably to Neighbor Joining, with the advantage of making
few or no false assertions about the topology of the model tree; guar­
antees against false positives can be controlled as a parameter by the
user.

Source: Anderson, Richard - Department of Computer Science and Engineering, University of Washington at Seattle

Collections: Computer Technologies and Information Sciences