Current browse context:
cs.LG
Change to browse by:
References & Citations
Computer Science > Machine Learning
Title: EvoVGM: a Deep Variational Generative Model for Evolutionary Parameter Estimation
(Submitted on 25 May 2022 (v1), last revised 1 Jul 2022 (this version, v2))
Abstract: Most evolutionary-oriented deep generative models do not explicitly consider the underlying evolutionary dynamics of biological sequences as it is performed within the Bayesian phylogenetic inference framework. In this study, we propose a method for a deep variational Bayesian generative model (EvoVGM) that jointly approximates the true posterior of local evolutionary parameters and generates sequence alignments. Moreover, it is instantiated and tuned for continuous-time Markov chain substitution models such as JC69, K80 and GTR. We train the model via a low-variance stochastic estimator and a gradient ascent algorithm. Here, we analyze the consistency and effectiveness of EvoVGM on synthetic sequence alignments simulated with several evolutionary scenarios and different sizes. Finally, we highlight the robustness of a fine-tuned EvoVGM model using a sequence alignment of gene S of coronaviruses.
Submission history
From: Amine Remita [view email][v1] Wed, 25 May 2022 20:08:10 GMT (33873kb,D)
[v2] Fri, 1 Jul 2022 01:28:55 GMT (15926kb,D)
Link back to: arXiv, form interface, contact.