A Variational Approach to Bayesian Phylogenetic Inference

被引:0
作者
Zhang, Cheng [1 ,2 ]
Matsen IV, Frederick A. [3 ,4 ]
机构
[1] Peking Univ, Sch Math Sci, Beijing 100871, Peoples R China
[2] Peking Univ, Ctr Stat Sci, Beijing 100871, Peoples R China
[3] Univ Washington, Dept Genome Sci, Seattle, WA 98195 USA
[4] Univ Washington, Dept Stat, Seattle, WA 98195 USA
基金
中国国家自然科学基金;
关键词
Bayesian phylogenetic inference; variational inference; subsplit Bayesian networks; structured amortization; POPULATION-DYNAMICS; DNA-SEQUENCES; F-DIVERGENCE; LIKELIHOOD; MODEL; DISTRIBUTIONS; EXPLORATION; PROPOSALS; EVOLUTION; HISTORY;
D O I
暂无
中图分类号
TP [自动化技术、计算机技术];
学科分类号
0812 ;
摘要
Bayesian phylogenetic inference is currently done via Markov chain Monte Carlo (MCMC) with simple proposal mechanisms. This hinders exploration efficiency and often requires long runs to deliver accurate posterior estimates. In this paper, we present an alternative approach: a variational framework for Bayesian phylogenetic analysis. We propose combining subsplit Bayesian networks, an expressive graphical model for tree topology distributions, and a structured amortization of the branch lengths over tree topologies for a suitable variational family of distributions. We train the variational approximation via stochastic gradient ascent and adopt gradient estimators for continuous and discrete variational parameters separately to deal with the composite latent space of phylogenetic models. We show that our variational approach provides competitive performance to MCMC, while requiring much fewer (though more costly) iterations due to a more efficient exploration mechanism enabled by variational inference. Experiments on a benchmark of challenging real data Bayesian phylogenetic inference problems demonstrate the effectiveness and efficiency of our methods.
引用
收藏
页码:1 / 56
页数:56
相关论文
共 107 条
[1]   α-Divergence Is Unique, Belonging to Both f-Divergence and Bregman Divergence Classes [J].
Amari, Shun-Ichi .
IEEE TRANSACTIONS ON INFORMATION THEORY, 2009, 55 (11) :4925-4931
[2]  
AMARI SI, 1985, LECT NOTES STAT, V28
[3]   BEAGLE 3: Improved Performance, Scaling, and Usability for a High-Performance Computing Library for Statistical Phylogenetics [J].
Ayres, Daniel ;
Cummings, Michael ;
Baele, Guy ;
Darling, Aaron ;
Lewis, Paul ;
Swofford, David ;
Huelsenbeck, John ;
Lemey, Philippe ;
Rambaut, Andrew ;
Suchard, Marc .
SYSTEMATIC BIOLOGY, 2019, 68 (06) :1052-1061
[4]   Genealogical Working Distributions for Bayesian Model Testing with Phylogenetic Uncertainty [J].
Baele, Guy ;
Lemey, Philippe ;
Suchard, Marc A. .
SYSTEMATIC BIOLOGY, 2016, 65 (02) :250-264
[5]  
Barber D, 1999, ADV NEUR IN, V11, P183
[6]  
Bishop C. M., 2006, Pattern Recognition and Machine Learning
[7]  
Bishop CM, 1998, ADV NEUR IN, V10, P416
[8]   Variational Inference: A Review for Statisticians [J].
Blei, David M. ;
Kucukelbir, Alp ;
McAuliffe, Jon D. .
JOURNAL OF THE AMERICAN STATISTICAL ASSOCIATION, 2017, 112 (518) :859-877
[9]  
Bornschein Jorg, 2015, P INT C LEARN REPR I
[10]   Ultrafast Approximation for Phylogenetic Bootstrap [J].
Bui Quang Minh ;
Minh Anh Thi Nguyen ;
von Haeseler, Arndt .
MOLECULAR BIOLOGY AND EVOLUTION, 2013, 30 (05) :1188-1195