Treelink: data integration, clustering and visualization of phylogenetic trees
© Allende et al. 2015
Received: 25 July 2015
Accepted: 15 December 2015
Published: 29 December 2015
Phylogenetic trees are central to a wide range of biological studies. In many of these studies, tree nodes need to be associated with a variety of attributes. For example, in studies concerned with viral relationships, tree nodes are associated with epidemiological information, such as location, age and subtype. Gene trees used in comparative genomics are usually linked with taxonomic information, such as functional annotations and events. A wide variety of tree visualization and annotation tools have been developed in the past, however none of them are intended for an integrative and comparative analysis.
Treelink is a platform-independent software for linking datasets and sequence files to phylogenetic trees. The application allows an automated integration of datasets to trees for operations such as classifying a tree based on a field or showing the distribution of selected data attributes in branches and leafs. Genomic and proteonomic sequences can also be linked to the tree and extracted from internal and external nodes. A novel clustering algorithm to simplify trees and display the most divergent clades was also developed, where validation can be achieved using the data integration and classification function. Integrated geographical information allows ancestral character reconstruction for phylogeographic plotting based on parsimony and likelihood algorithms.
Our software can successfully integrate phylogenetic trees with different data sources, and perform operations to differentiate and visualize those differences within a tree. File support includes the most popular formats such as newick and csv. Exporting visualizations as images, cluster outputs and genomic sequences is supported. Treelink is available as a web and desktop application at http://www.treelinkapp.com.
Phylogenetic trees are increasingly used to visualize comparative information in an evolutionary context , e.g. comparing strands of viruses and genes within different clades to find new ways of classifying and differentiating groups of leafs within a tree. A wide variety of visualization and annotation tools have been developed in the past [2–4] however none of them are intended for integrative and comparative analysis.
Information that can be interpreted in a phylogenetic context is growing rapidly, creating a continuous need to find new ways to integrate, process and deliver this new information. Furthermore, there is also a demand to explore, compare, display and interpret trees using information not directly contained in these trees, such as taxonomy, geography, and traits, among others .
Treelink was designed with those needs in mind, enabling an easy and automated procedure that links data sources to nodes and topologies, which in turn allow the construction of comparative representations that can be explored and extracted.
Representation of large phylogenies and clustering of nodes has also proven to be difficult in epidemiology and evolutionary research , where large and complex trees are used for exploration and pattern analysis. A novel clustering algorithm was developed that groups and divides clades and sequences within a tree topology based on their divergence measure, resulting in characteristic and representative sets that simplify phylogenies and reveal similarity.
The main design directives behind Treelink were: (i) Ease of use (ii) Automation (iii) Fast Performance (iv) Scalability (v) Aesthetics.
The application’s first step is to provide automated data integration and visualization for additional renderings of the tree. The main features include:
Data integration from standard datasets is executed by linking the leaf label to a key in the dataset. Once the integration process is finished, the aggregated information is displayed by hovering over the nodes. The corresponding table can also be shown for additional operations.
Classification and cross-reference
Sequence extraction and visualization
Sequence information in FASTA format can also be linked to the selected phylogeny for extraction and visualization purposes. By selecting leafs and internal nodes, the user can visualize or export the subset of genomic or proteonomic sequences into a new file.
Clustering and tree simplification
Treelink also includes a clustering and tree simplification component called TreeClus, based on a novel algorithm that divides and clusters phylogenetic topologies based on a dissimilarity measure calculated from the evolutionary distance of the branches (for more details, (see Additional file 1)). This function can be used for detecting characteristic subgroups within a given topology for differential analysis in evolution, e.g. discerning subtypes in viral phylogenies  and classifying taxa.
Ancestral reconstruction and phylogeography
Phylogeographic inference from tree topologies and datasets is implemented in the TreeMap component, through ancestral character state reconstruction  for different evolutionary models: Linear parsimony with a delayed and accelerated transformation rule, and a likelihood approach , for cases assuming some version of Brownian motion. The resulting tree and states can then be plotted on a map, where the movement of states is animated proportional to evolutionary time.
Visualizations of attribute searches or a classification by a field
FASTA files from a subtree or leaf
Tabulated files of the sequences and their cluster
Clustered Tree Visualizations
Ancestral Reconstruction Visualizations
A major point of difference between Treelink and alternative software (e.g., iTol, TreeDyn) [2–4, 10] is its automated data integration process, that doesn’t require formatting or pre-processing on datasets to fit input requirements. Other existing tools require manual annotation of meta-data to associate or attach information to selected tree elements. Treelink overcomes these requirements by using standard dataset formats as an integration source, relieving the user of tasks like manual annotations on the leafs and permits integrating associated data directly from the sources of data collection, given that csv is a popular export format of sql-based databases, excel and other spreadsheets.
Another upside includes the amount of fields that can be linked to the tree, allowing up to 9 different fields to be integrated. Performance and flexibility distinctions include the ability to navigate and interact with the tree by selecting subtrees and creating diverse visualizations almost instantaneously after the initial data integration step with the same data source.
The design based on simplicity requires little training or knowledge of the tool, labels are annotated automatically with corresponding legends and attributes can be search directly into the tree. An short interactive tutorial found on the site gives a quick overview of the main functions and use of the tool. A short manual is also included for additional knowledge on the limitations and capabilities of Treelink.
Treelink is a comprehensive open-source solution for rendering and integrating datasets into phylogenetic trees. It supports all operating systems and has an HTML5 version for the widest possible audience. File support includes the most popular formats like newick and csv, generating interactive trees that incorporate analytics in scalable vector images. It can also be used as a library to extend and complement with additional functions and methods, and can be easily integrated into existing web applications.
Availability and requirements
Thanks to Héctor Allende Olivares for helping with the mathematical explanations of the clustering algorithm.
Open Access This article is distributed under the terms of the Creative Commons Attribution 4.0 International License (http://creativecommons.org/licenses/by/4.0/), which permits unrestricted use, distribution, and reproduction in any medium, provided you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons license, and indicate if changes were made. The Creative Commons Public Domain Dedication waiver (http://creativecommons.org/publicdomain/zero/1.0/) applies to the data made available in this article, unless otherwise stated.
- Davies T, Fritz S, Grenyer R, Orme C, Bielby JEA. Phylogenetic trees and the future of mammalian biodiversity. PNAS. 2008; 105:11556–63.View ArticlePubMedPubMed CentralGoogle Scholar
- Letunic I, Bork P. Interactive tree of life (itol): an online tool for phylogenetic tree display and annotation. Bioinformatics. 2007; 23:127–8.View ArticlePubMedGoogle Scholar
- Page R. Visualizing phylogenetic trees using treeview. Curr Protocol Bioinforma. 2002; 6(2):Unit 6.2.Google Scholar
- Tao Y, Liu Y, Friedman C, Lussier Y. Information visualization techniques in bioinformatics during the postgenomic era. Drug Discovery Today: BIOSILICO. 2004; 2(6):237–45.View ArticlePubMedPubMed CentralGoogle Scholar
- Lott P, Mundry M, Sassenberg C, Lorkowski S, Fuellen G. Simplifying gene trees for easier comprehension. BMC Bioinformatics. 2006;7(231).Google Scholar
- Pasquier C, Millot N, Njouom R, Sandres K, Cazabat M, Puel J, et al. Hiv-1 subtyping using phylogenetic analysis of pol gene sequences. J Virol Methods. 2001; 94(1–2):45–54.View ArticlePubMedGoogle Scholar
- Cunningham CW, Omland KE, Oakley TH. Reconstructing ancestral character states: a critical reappraisal. Trends Ecol Evol. 1998; 13(9):361–6.View ArticlePubMedGoogle Scholar
- Lewis P. A likelihood approach to estimating phylogeny from discrete morphological character data. Syst Biol. 2001; 50(6):913–25.View ArticlePubMedGoogle Scholar
- Chevenet F, Brun C, Bañuls A, Jacq B, Christen R. Treedyn: towards dynamic graphics and annotations for analyses of trees. BMC Bioinformatics. 2006;7(439).Google Scholar
- Holmes E, Nee S, Rambaut A, Garnett G, PH H. Revealing the history of infectious disease epidemics through phylogenetic trees. Phil Trans Biol Sci. 1995; 349(1327):33–40.View ArticleGoogle Scholar
- Beheregaray L. Twenty years of phylogeography: the state of the field and the challenges for the southern hemisphere. Mol Ecol. 2008; 17:3754–74.PubMedGoogle Scholar
- Swofford DL, Maddison WP. Reconstructing ancestral character states under wagner parsimony. Math Biosci. 1987; 87:199–229.View ArticleGoogle Scholar