D. Huson, S. Nettles, K. Rice, T. Warnow, and S. Yooseph. The hybrid tree reconstruction method. ACM J. Experimental Algorithmics, 4(5), 1999. http://www.jea.acm.org/1999/HusonHybrid/.  Home/Search   Document Details and Download   Summary   Related Articles   Check

.... that are close to the true evolutionary distances will in general be more useful for evolutionary tree reconstruction than edit distances, because edit distances underestimate true evolutionary distances, and this underestimation can be very significant as the number of rearrangements increases [7, 20]. There are two criteria for evaluating a t.e.d. estimator: how close the estimated distances are to the true evolutionary distance between two genomes, and how accurate the inferred trees are when a distance based method (e.g. neighbor joining) is used in conjunction with these distances. The ....

....by the linear time algorithm for minimum inversion distances given in [2] between any two genomes in the input. Distance matrices with some normalized edit distances close to 1 are said to be saturated , and the recovery of accurate trees from such datasets is considered to be very dicult [7]. The y axis is the false negative rate (i.e. the proportion of missing edges) False negative rates of less than 5 are excellent, but false negative rates of up to 10 can be tolerated. We use NJ(D) to denote the tree returned by NJ using distance D. Note that except for NJ(EDE) the relative ....

D. Huson, S. Nettles, K. Rice, T. Warnow, and S. Yooseph. The hybrid tree reconstruction method. J. Experimental Algorithmics, 4:178-189, 1999. http://www.jea.acm.org/.

....any edge in the tree T 0 . This tree always exists, since the star tree trivially satis es the constraint on any set of quartets. The Q tree is unique and can be constructed in polynomial time; by design, however, the Q method is conservative and generally produces very unresolved trees [11]. Quartet Cleaning (QC) methods [2, 4, 13] have explicit bounds on the number of quartet errors around each reconstructed edge e. These error bounds have the form m p q e , where q e is the number of quartets around edge e and m is a small constant. Thus, the Q method is a cleaning method ....

....known upper bounds on the convergence rates of NJ and the Q method. Surprisingly, these are identical [1, 6] although experimental studies strongly suggest that NJ obtains accurate reconstructions of trees from shorter sequences than Q throughout the parameter space of Jukes Cantor trees [11]. Theorem 3.1. Let f; g; 0 be arbitrary constants with f g. Denote by B(S) the tree reconstructed on S by the Q method and by NJ(S) the tree reconstructed by NJ. There is a constant c 0 such that, for all Jukes Cantor trees (T ; f e g) on n leaves with 0 f e g 1 for all e 2 ....

[Article contains additional citation context not shown here]

Huson, D., Nettles, S., Rice, K., Warnow, T., and Yooseph, S., \The hybrid tree reconstruction method," ACM J. Experimental Algorithms 4, 5 (1999), www.jea.acm.org/1999/HusonHybrid/.

No context found.

D. Huson, S. Nettles, K. Rice, T. Warnow, and S. Yooseph. The hybrid tree reconstruction method. J. Experimental Algorithmics, 4:178--189, 1999. http://www.jea.acm.org/.

.... very fast (polynomial time and fast in practice) While the experimental evidence is not yet definitive, the best distance based methods appear less accurate than the better heuristics for maximum parsimony and maximum likelihood, at least on large trees with high rates of evolution (see, e.g. [13]) 3 Algorithmic and Experimental Challenges 3.1 Designing for Speed Because both parsimony and likelihood based approaches involve NP hard optimization problems and because poor approximations may lead to biologically incorrect conclusions, developing efficient exact algorithms is a major ....

D. Huson, S. Nettles, K. Rice, T. Warnow, and S. Yooseph. Hybrid tree reconstruction methods. ACM Journal of Experimental Algorithmics, 4(5), 1999. Online at www.jea.acm.org/1999/HusonHybrid/.

....the input set of quartets Q. This tree always exists, since the star tree trivially satisfies the constraint on any set of quartets. The Q # tree is unique and can be constructed in polynomial time. By design, however, the Q # method is conservative and generally produces very unresolved trees [13]. Quartet Cleaning (QC) methods [2, 4, 15] have explicit bounds on the number of quartet errors around each reconstructed edge e. As with the Q # method, we start with an input set of quartets Q. For each edge e in the true tree T , let q e be the number of quartets that cross e (that is, all ....

....the known upper bounds on the convergence rates of NJ and the Q # method. Surprisingly, these are identical [1, 7] although experimental studies strongly suggest that NJ obtains accurate reconstructions of trees from shorter sequences than Q # throughout the parameter space of Jukes Cantor trees [13]. We then discuss upper bounds on quartet cleaning. Experimental results illustrating the tightness of the upper bounds can be found in [13, 14] and Section 5.7. Theorem 3.1. Let f, g, # 0 be arbitrary constants with f g. Denote by Q # (S) the tree reconstructed on S by the Q # method and by ....

[Article contains additional citation context not shown here]

Huson, D., Nettles, S., Rice, K., Warnow, T., and Yooseph, S., "The hybrid tree reconstruction method," ACM J. Experimental Algorithms 4, 5 (

No context found.

D. Huson, S. Nettles, K. Rice, T. Warnow, and S. Yooseph. The hybrid tree reconstruction method. J. Experimental Algorithmics, 4:178 189, 1999. http://www.jea.acm.org/.

No context found.

D. Huson, S. Nettles, K. Rice, T. Warnow, and S. Yooseph. The hybrid tree reconstruction method. J. Experimental Algorithmics, 4:178--189, 1999. http://www.jea.acm.org/.

....the parameter space in which MPBE or BPAnalysis outperform neighbor joining in topological accuracy, there is also an argument to be made in favor of the fast neighbor joining method. Indeed, we conjecture that such regions do exist (as other studies based upon biomolecular sequence evolution show [24, 18]) and our research will also seek to determine whether this is true for gene order data as well. 6 Recommendations Given the rapid increase in the availability of complete genome sequences, the current limitation in reconstructing phylogenies from gene order data for datasets containing many ....

D. Huson, S. Nettles, K. Rice, T. Warnow, and S. Yooseph. "The Hybrid tree reconstruction method." To appear, The Journal of Experimental Algorithms, special issue for selected papers from The Workshop on Algorithms Engineering, Saarbrucken, Germany, 1998.

.... This saturation value depends upon the number of genes and is bounded from above by n, the maximum distance between any two genomes on n genes [18] Reconstructing trees from saturated datasets is difficult because of the seeming randomness in the data this is well understood for gene sequences [15, 16, 28], so it is no surprise that it applies to gene rearrangements as well. Given two genomes G and G 0 on the same set of genes, a breakpoint in G is defined as an ordered pair of genes (g i ; g j ) such that g i and g j appear consecutively in that order in G, but neither (g i ; g j ) nor ( g j ; ....

....might be obtained by correcting distances for genome rearrangements. Furthermore, there is both theoretical and empirical evidence that the trees reconstructed by most distance methods, including NJ, degrade significantly (in terms of topological accuracy) under high rates of evolution [1, 15]. For these reasons, we have given a lot of attention to providing improved estimates of inversion distances. 5.2 IEBP a theoretical correction We have developed a general analytical technique for estimating the expected number of breakpoints produced by a sequence of random events in the ....

Huson, D., Nettles, S., Rice, K., Warnow, T., & Yooseph, S., "Hybrid tree reconstruction methods," ACM J. Experimental Algorithmics 4, 5 (1999), www.jea.acm.org/1999/HusonHybrid/.

....and hence help us identify alternative hypotheses. The task remains to identify regions of the parameter space in which MPBE or BPAnalysis outperform neighbor joining in topological accuracy. We conjecture that such regions do exist (as other studies based upon biomolecular sequence evolution show [24, 16]) Given the rapid increase in the availability of complete genome sequences, the current limitation in reconstructing phylogenies from gene order data for datasets containing many genomes or genes is of major concern. Until improved methods are developed, we recommend that phylogenetic analyses ....

D. Huson, S. Nettles, K. Rice, T. Warnow, and S. Yooseph. "The Hybrid tree reconstruction method." To appear, The Journal of Experimental Algorithms, special issue for selected papers from The Workshop on Algorithms Engineering, Saarbrucken, Germany, 1998.

No context found.

Huson, D., Nettles, S., Rice, K., Warnow, T., and Yooseph, S., \The hybrid tree reconstruction method," ACM J. Experimental Algorithmics 4, 5 (1999), www.jea.acm.org/1999/HusonHybrid/.

....and hence help us identify alternative hypotheses. The task remains to identify regions of the parameter space in which MPBE or BPAnalysis outperform neighbor joining in topological accuracy. We conjecture that such regions do exist (as other studies based upon biomolecular sequence evolution show [24, 16]) Given the rapid increase in the availability of complete genome sequences, the current limitation in reconstructing phylogenies from gene order data for datasets containing many genomes or genes is of major concern. Until improved methods are developed, we recommend that phylogenetic analyses ....

D. Huson, S. Nettles, K. Rice, T. Warnow, and S. Yooseph. "The Hybrid tree reconstruction method." To appear, The Journal of Experimental Algorithms, special issue for selected papers from The Workshop on Algorithms Engineering, Saarbrucken, Germany, 1998.

.... methods for solving parsimony continue to be a major technique used by systematic biologists, even for large trees (see, for example, of a 500 taxon data set) despite its computational difficulties and its inconsistency under some models of evolution [30, 17] Furthermore, both experimental [34, 27] and analytical studies [15, 5, 3, 16] suggest that the actual sequence length that suffices for the various promising distance based methods to obtain (with high probability) an accurate topology estimation may be exceedingly large if the true tree contains significant divergence, even under the ....

....divergence. On the other hand, heuristic approaches for the maximum parsimony problem have performed very well in experimental studies, and often perform significantly better than the best polynomial time distance based methods, such as neighbor joining [35] in reconstructing very large trees [25, 34, 27]. For this reason, many important data sets are analyzed in biology using heuristic parsimony searches. Unfortunately these searches can last for months or longer (see [33] without necessarily finding the most parsimonious tree(s) This is partly due to the fact that parsimony is NP hard to solve ....

[Article contains additional citation context not shown here]

D. Huson, S. Nettles, K. Rice, T. Warnow, and S. Yooseph, Hybrid Tree Reconstruction Methods, 2nd Workshop on Algorithm Engineering (WAE'98), Saarbrucken, Germany, 1998.

No context found.

D. Huson, S. Nettles, K. Rice, T. Warnow, and S. Yooseph. The hybrid tree reconstruction method. ACM J. Experimental Algorithmics, 4(5), 1999. http://www.jea.acm.org/1999/HusonHybrid/.

Online articles have much greater impact   More about CiteSeer.IST   Add search form to your site   Submit documents   Feedback  

CiteSeer.IST - Copyright Penn State and NEC