F. Chevenet

Research engineer in computational biology, bioinformatics, in the field of information visualization and phylogeny.MIVEGEC: Maladies Infectieuses et Vecteurs Ecologie, Génétique, Évolution et ContrôleIRD Montpellier911 Avenue Agropolis BP 64501 34394 Montpellier cedex 5 - Francefrancois.chevenet@ird.frMAB: Méthodes et Algorithmes pour la BioinformatiqueLaboratoire d'Informatique Robotique et Microélectronique de Montpellier (LIRMM)161 rue Ada 34095 Montpellier Cedex 5 - Francefrancois.chevenet@lirmm.fr

EvoLaps: a web interface to visualize continuous phylogeographic reconstructions

Web site: www.evolaps.orgChevenet F, Fargette D, Guindon S, Bañuls AL submitted to BMC Bioinformatics 2021
Background: phylogeographic reconstructions serve as a basis to understand the spread and evolution of pathogens. Visualization of these reconstructions often lead to complex graphical representations which are difficult to interpret.
Result: we present EvoLaps, a user-friendly web interface to visualize phylogeographic reconstructions based on the analysis of latitude/longitude coordinates with various clustering levels. EvoLaps also produces transition diagrams that provide concise and easy to interpret summaries of phylogeographic reconstructions.
Conclusion: The main contribution of EvoLaps is to assemble known numerical and graphical methods/tools into a user-friendly interface dedicated to the visualization and edition of evolutionary scenarios based on continuous phylogeographic reconstructions. EvoLaps is freely usable at www.evolaps.org.
Keywords: phylogeography, ancestral character states, evolutionary scenario, data visualization

EvoLaps interface. The interface is composed of (a) a left panel gathering three toolboxes, corresponding to the steps of the analysis: ‘data’, ‘clustering’ and ‘edition’; (b) a main view (right panel), subdivided into three components during an analysis: (b1) a geographic map displaying the spatial distribution of the samples, clusters and the resulting paths (spread between the clusters), (b2) the phylogenetic tree displaying the evolutionary relationships among the samples and (b3) a transition diagram summarizing the transitions between the clusters. These three components are inter-connected and they share the same color code based on the current set of clusters defined by the user during the analysis.

PastView: a user-friendly interface to explore ancestral scenarios

Web site: www.pastview.orgChevenet F, Castel G, Jousselin E, Gascuel O. BMC Evolutionary Biology 2019
Phylogenetic trees are often combined with extrinsic traits to reconstruct evolutionary scenarios in various research fields, including phylogeography, molecular epidemiology, and ecology. Analysis starts with the computation of ancestral annotations from extrinsic traits (discrete variables such as geographic origin, resistance to a treatment, and life history traits), associated with the sampled sequences used to build the phylogenetic tree. Analysis continues with the study of the evolutionary changes from the tree root to its tips to characterize evolutionary scenarios such as the spread of a disease, the dynamic of a drug resistance, or shifts in ecological habitats. Difficulties may arise when interpreting the output. First, there is a diversity of optimization criteria for computing trees and ancestral annotations (parsimony, maximum likelihood, Bayesian) and each of them involves various models of evolution; these methods may all yield different results that need to be compared. Second, the tree size and complexity of the annotations can be an inconvenience with respect to computation time or tedious graphical analysis. Third, although probabilistic models give more accurate results than other methods, if some ancestral characters do not stand out clearly from others (i.e. they are much more likely), they may produce a combinatorial explosion of potential evolutionary scenarios. To the best of our knowledge, there are no tools to help with the comparison of different sets of ancestral annotations and find common patterns across multiple evolutionary scenarios from a beam of transitions. We thus present PastView, a new editor with a user-friendly interface that includes numerical and graphical features to help users in this context. PastView is available publicly as a standalone software.

Output of a PastView analysis for the study of HIV-1A epidemiological history in Albania (Salemi et al., 2008). The countries associated with the tree sequences are used to compute ancestral areas by three methods: maximum a posteriori (MAP), the joint most likely scenario, and parsimony (DELTRAN). (a) A country color-code is used to color nodes and branches if their associated ancestral annotations are the same for the three methods; if not, bubbles are displayed. (b) A filter (threshold based on MAP probability minus 40% of its value) is used to display pie charts of posteriors for ambiguous nodes. (c), (d) Tree-like representations of transitions (MAP and joint inferences, respectively); numbers indicate the numbers of identical transitions having the same ancestor in the transition maps type 1. (e) The transition query ‘* Albania’ highlights the tree pathways from the root to Albania; it displays the two distinct transitions from Greece to Albania.

PhyloType: Searching for virus phylotypes

Web site: www.phylotype.orgChevenet F, Jung M, Peeters M, de Oliveira T, Gascuel O. Bioinformatics 2013
Very large phylogenies are being built today in order to study virus evolution, trace the origin of epidemics, establish the mode of transmission, and survey the appearance of drug resistance. However, no tool is available to quickly inspect these phylogenies and combine them with extrinsic traits (e.g. geographic location, risk group, presence of a given resistance mutation), seeking to extract strain groups of specific interest or requiring surveillance. We propose a new method for obtaining such groups, which we call phylotypes, from a phylogeny having taxa (strains) annotated with extrinsic traits. Phylotypes are subsets of taxa with close phylogenetic relationships and common trait values. The method combines ancestral trait reconstruction using parsimony, with combinatorial and numerical criteria measuring tree shape characteristics and the diversity and separation of the potential phylotypes. A shuffling procedure is used to assess the statistical significance of phylotypes. All algorithms have linear time complexity. This results in very low computing times.

Tree graphics obtained in the study of the epidemiological history of HIV-1A in Albania (Salemi et al., 2008). (a) Phylogenetic tree in “background” format: selected phylotypes and their strains are colored; colored regions comprise all (uniquely annotated) nodes on the path from the phylotype root to the phylotype members; not colored (black) strains do not belong to any phylotype; the root node identifiers of phylotypes are provided, to be used in conjunction with the detailed table (Table 1). (b) Phylotype map, summarizing the information contained in phylogenetic tree (a); circle surface is proportional to the Size value (number of members) of the phylotype

SylvX: a viewer for phylogenetic tree reconciliations

Web site: www.sylvx.orgChevenet F , Doyon JP, Scornavacca C, Jacox E, Jousselin E, Berry V Bioinformatics 2015
Reconciliation methods aim at recovering the evolutionary processes that shaped the history of a given gene family including events such as duplications, transfers and losses by comparing the discrepancies between the topologies of the associated gene and species trees. These methods are also used in the framework of host/parasite studies to recover co-diversification scenarios including co-speciation events, host-switches and extinctions. These evolutionary processes can be graphically represented as nested trees. These interconnected graphs can be visually messy and hard to interpret, and despite the fact that reconciliations are increasingly used, there is a shortage of tools dedicated to their graphical management. Here we present SylvX, a reconciliation viewer which implements classical phylogenetic graphic operators (swapping, highlighting, etc.) and new methods to ease interpretation and comparison of reconciliations (multiple maps, moving, shrinking sub-reconciliations). SylvX is an open source, cross-platform, standalone editor available for Windows and Unix-like systems including OS X. It is publicly available at www.sylvx.org. Contact: francois.chevenet@ird.fr

Screenshot of a SylvX session. The left side of the interface gathers the Map toolbox, the Reconciliation(s) toolbox and the Species tree toolbox. The right side stacks maps. Duplications are represented by blue squares, losses by red crosses and transfers by arrows (variable colors). The ‘host-parasite’ dataset used in the figure consists of the Ficus /fig wasp dataset by Cruaud et al. (2012) . Two maps are in use: the one on the left displays a species tree with a superimposed reconciliation having a color coding based on transfers. The second map displays a reconciliation for the same gene but with different costs for evolutionary events

ScripTree: scripting phylogenetic graphics

Web site: www.scriptree.orgChevenet F, Croce O, Hebrard M, Christen R and Berry V Bioinformatics 2010
There is a large amount of tools for interactive display of phylogenetic trees. However, there is a shortage of tools for the automation of tree rendering. Scripting phylogenetic graphics would enable the saving of graphical analyses involving numerous and complex tree handling operations and would allow the automation of repetitive tasks. ScripTree is a tool intended to fill this gap. It is an interpreter to be used in batch mode. Phylogenetic graphics instructions are stored in a text file and performed in a sequential way. Such instructions are related to tree rendering as well as tree annotation. ScriptTree is written in Tcl/Tk making it a cross-platform application, e.g. suitable for Windows and Unix-like systems, including OS X. It can be used either as a standalone package or included in a bioinformatic pipeline and linked to a HTTP server.

Sample of a script (a) and image generated by its interpretation by ScripTree (b) for four gene trees of 19 virus species (Simon et al., 2005). ScripTree's input is (i) a tree file containing four newick strings; (ii) a script file with the commands; (iii) an annotation file with two variables: Genus and Capsid. The Genus variable stores taxonomic information related to two genera, Nucleopolyhedrovirus (NPV) and Granulovirus (GV). The Capsid variable codes for single (S) or multiple (M) virion nucleocapsids.

TreeDyn: towards dynamic graphics and annotations for analyses of trees

Web site: www.treedyn.orgPublication: Chevenet F, Brun C, Banuls AL, Jacq B and Christen R BMC Bioinformatics 2006
Many powerful tree editors are now available, but existing tree visualisation tools make little use of meta-information related to the entities under study such as taxonomic descriptions, geographic distribution or gene functions. This meta-information is useful for the analyses of trees and their publications, but can hardly be encoded within the tree itself (the so-called newick format). Consequently, a tedious manual analysis and post-processing of the tree's images is required. Particularly with large trees, multiple trees and multiple meta-information variables. TreeDyn links unique leaf labels to lists of variables/values pairs of annotations (meta-information), independently of the tree topologies, remaining fully compatible with the basic newick format. These relationships are used by dynamic graphics operators, information visualization methods like Projection, Localization, Labelization, Reflection allowing an interaction from annotations to trees, from trees to annotations and from trees to trees through annotations.

A functional classification tree for 602 yeast proteins computed with the PRODISTIN method. (a) The foundation for protein clustering. PRODISTIN classes are clustered according to the 'cellular role' of proteins only (pink), according to the 'functional category' of proteins only (blue), and according to both criteria (yellow). (b) Functional classification. PRODISTIN classes on the circular classification tree have been colored according to their corresponding 'cellular role'.

Page updated

Google Sites

Report abuse