Hybrid Genetic Algorithm and Lasso Test Approach for Inferring Well Supported Phylogenetic Trees Based on Subsets of Chloroplastic Core Genes

Abstract

Abstract. The amount of completely sequenced chloroplast genomes increases rapidly every day, leading to the possibility to build large scale phylogenetic trees of plant species. Considering a subset of close plant species defined according to their chloroplasts, the phylogenetic tree that can be inferred by their core genes is not necessarily well supported, due to the possible occurrence of "problematic" genes (i.e., homoplasy, incomplete lineage sorting, horizontal gene transfers, etc.) which may blur phylogenetic signal. However, a trustworthy phylogenetic tree can still be obtained if the number of problematic genes is low, the problem being to determine the largest subset of core genes that produces the best supported tree. To discard problematic genes and due to the overwhelming number of possible combinations, we propose an hybrid approach that embeds both genetic algorithms and statistical tests. Given a set of organisms, the result is a pipeline of many stages for the production of well supported phylogenetic trees. The proposal has been applied to different cases of plant families, leading to encouraging results for these families.

Binary Particle Swarm Optimization Versus Hybrid Genetic Algorithm for Inferring Well Supported Phylogenetic Trees

“…If so, we face an optimization problem: find the most supported tree using the largest subset of core genes. Obviously, a brute force approach investigating all possible combinations of genes is intractable, as it leads to 2 n phylogenetic tree inferences for a core genome of size n. To solve this optimization problem, we have formerly proposed in [1] a general pipeline detailed in Figure 1. In this pipeline, the stage of phylogenetic tree analysis mixes both genetic algorithm with LASSO tests in order to discover problematic genes.…”

“…In order to illustrate how to use the BPSO approach, we have considered the Rosales order, which has already been analyzed in [1] using a hybrid genetic algorithm and Lasso test approach. The Rosales order is constituted by 9 ingroup species and 1 outgroup (Mollissima), as described in Table 2.…”

“…At each iteration, the particle velocity is updated by taking into account its own best position and the best one considering the whole particle swarm (both identified according to the fitness value). It is influenced by constant weight factors as expressed in Equation (1). In this one, we have set φ 1 = c 1 · r 1 and φ 2 = c 2 · r 2 where c 1 = 1 and c 2 = 1, while r 1 , r 2 are random numbers belonging to [0.1,0.5], and w is the inertia weight that is computed based on the following formula: where I max represents the maximum number of iterations (or time step) and I cur is the current iteration.…”

“…A previous work [1] has proposed the use of an ad hoc Genetic Algorithm (GA) to solve the problem of finding the largest subset of core genes producing a phylogenetic tree as supported as possible. However, in some situations, this algorithm fails to solve the optimization problem due to a low convergence rate.…”

Systematic investigations of gene effects on both topologies and supports: An Echinococcus illustration

“…This is why phylogenetic analyses of animals are frequently performed by using a subset of mitochondrial coding sequences, 5 while chloroplasts are often considered for plant phylogenies. [6][7][8] For instance, in tetrapods and mammals, the best results have been obtained with nad5, nad4, nad2, cytb, and cox. 9 The optimal size and the content of this subset of mt genes have however signi¯cant consequences: a smaller subset of sequences necessitates a lower cost for DNA extraction and sequencing, while demanding a lower computation time.…”