IMGT/Phylogene documentation

Introduction
IMGT/PhyloGene sequence selection page
IMGT/PhyloGene selected sequences page
IMGT/PhyloGene distance matrix result page
IMGT/PhyloGene tree drawing page
IMGT/PhyloGene tree alignment page
IMGT/PhyloGene implementation and developments
IMGT index

Introduction

IMGT/PhyloGene [2] is part of IMGT, the international ImMunoGeneTics information system® http://www.imgt.org [3].

IMGT/PhyloGene, created by LIGM (Université Montpellier II, CNRS) is an IMGT tool for sequence analysis, on the Web since 10 February 2002.

IMGT/PhyloGene is an easy way to use tool for phylogenetic analysis of IMGT standardized reference sequences. IMGT/PhyloGene is an on line software package which computes and draws phylogenetic trees for immunoglobulin (IG) and T-cell receptor (TR) V-REGION nucleotide sequences.

IMGT/PhyloGene consists of 5 complementary tools:

A selection tool which allows to interactively and progressively select the IG and TR sequences to analyse.
An evolutionary distance calculation tool which computes a distance matrix [4] from the selected sequences.
A tree building tool which uses the Neighbor-Joining algorithm [5] to build a phylogenetic tree from the previously calculated distance matrix.
A tree drawing tool which provides a graphical representation of a rooted phylogenetic tree.
A label subsequences viewing tool which displays the subsequences corresponding to V-REGION related labels at the tips of a phylogenetic tree.

IMGT/PhyloGene sequence selection page

The nucleotide sequences available for selection in IMGT/PhyloGene are IG and TR V-REGIONs and C-DOMAINs, which follow the IMGT unique numbering (IMGT unique numbering for all IG and TR V-REGIONs of all species, IMGT unique numbering for C-DOMAIN and C-LIKE-DOMAIN ).
V-REGION and C-DOMAIN sequences are extracted from IMGT/GENE-DB. Note that only the first allele (allele *01) of each gene is available.
Three popup menus allow to select the species, the group and the subgroup(s), while one multiple selection list allows to select precisely the genes to analyse.

By default, the "Homo sapiens" "IGHV" genes are displayed. The species or the group can immediatly be changed by using the popup menus, and the page will automatically reload itself with the right information for the subgroups and genes (for example, select the "Mus musculus" "IGKV" group, and the "IGKV" subgroups and genes will be made available). Gene names belonging to the selected species, group, and subgroup(s) appear in a multiple selection list on the right side of the page. At first they all appear selected (in black), but they can be unselected if some of them are not required (press on "Ctrl"). Conversely, to select several genes, press "Ctrl" and "Alt" simultaneously when doing the selection.
Then, pressing the "Add to selection" button adds the selected genes to your selection (you may think of your selection as a "bag of sequences", thatis progressively filled with the sequences you wish to analyse).
Note that the number of sequences contained in your selection is indicated in red at the top of the page. Pressing the "Reset" button will empty your selection.
Using the text area at the bottom of the page, you can also paste your own sequences to add them to the analysis. These sequences must be in FASTA format, and must follow closely the IMGT unique numbering for V-REGION or C-DOMAIN. It is very important to respect both the maximum length (in number of nucleotides) and the gaps positions according to the IMGT numbering:
- for V-REGION: Definition of the FR-IMGT and CDR-IMGT regions. IMGT/V-QUEST can be used to help in positioning the gaps.
- for C-DOMAIN: Rules for the gap and additional positions in C-DOMAIN and C-LIKE-DOMAIN
Partial sequences should be completed with gaps.

Here are examples of sequences in the required format :

>X60503|IGHV1-18*02|Homo sapiens
CAGGTTCAGCTGGTGCAGTCTGGAGCT...GAGGTGAAGAAGCCTGGGGCCTCAGTGAAGGTCTCCTGCAAGGCTTCTG
GTTACACCTTTACCAGCTATGGT............ATCAGCTGGGTGCGACAGGCCCCTGGACAAGGGCTTGAGTGGAT
GGGATGGATCAGCGCTTACAATGGTAACACA......AACTATGCACAGAAGCTCCAG...GGCAGAGTCACCATGACC
ACAGACACATCCACGAGCACAGCCTACATGGAGCTGAGGAGCCTAAGATCTGACGACACGGCC

>X60503|IGHV1-18*02|Homo sapiens (partial)
..........................................................AGGTCTCCTGCAAGGCTTCTG
GTTACACCTTTACCAGCTATGGT............ATCAGCTGGGTGCGACAGGCCCCTGGACAAGGGCTTGAGTGGAT
GGGATGGATCAGCGCTTACAATGGTAACACA......AACTATGCACAGAAGCTCCAG...GGCAGAGTCACCATGACC
ACAGACACATCCACGAGCACAGCCTACATGGAGCTGAGGAGCCTAAGATCTGACGACACGGCC

>K01316|IGHG4*01_CH1|Homo sapiens
.............CTTCCACCAAGGGCCCATCCGTCTTCCCCCTGGCGCCCTGCTCCAGGAGCACCTCC.........G
AGAGCACAGCCGCCCTGGGCTGCCTGGTCAAGGACTACTTCCCC......GAACCGGTGACGGTGTCGTGGAACTCAGG
CGCCCTGACCAGC............GGCGTGCACACCTTCCCGGCTGTCCTACAGTCCTCA..................
GGACTCTACTCCCTCAGCAGCGTGGTGACCGTGCCCTCCAGCAGCTTG.........GGCACGAAGACCTACACCTGCA
ACGTAGATCACAAGCCC......AGCAACACCAAGGTGGACAAGAGAGTT

>X51755|IGLC1*01_CH1|Homo sapiens
..........GTCAGCCCAAGGCCAACCCCACTGTCACTCTGTTCCCGCCCTCCTCTGAGGAGCTCCAA.........G
CCAACAAGGCCACACTAGTGTGTCTGATCAGTGACTTCTACCCG......GGAGCTGTGACAGTGGCCTGGAAGGCAGA
TGGCAGCCCCGTCAAGGCG......GGAGTGGAGACCACCAAACCCTCCAAACAGAGCAAC..................
AACAAGTACGCGGCCAGCAGCTACCTGAGCCTGACGCCCGAGCAGTGG......AAGTCCCACAGAAGCTACAGCTGCC
AGGTCACGCATGAA............GGGAGCACCGTGGAGAAGACAGTGGCCCCTACAGAATGTTCA

Here is a screenshot of the IMGT/PhyloGene sequence selection page after selection of 8 sequences, and before selecting additional sequences ("Mus musculus" and "IGHV group"):

IMGT/PhyloGene selected sequences page

This page shows the selected sequences. The IMGT unique numbering allows to get an automatic alignement of the sequences. The delimitations of the FR-IMGT and CDR-IMGT are also shown.
Note that the CDR-IMGT will be removed for the phylogenetic analysis by default. Indeed, the various lengths do not allow a meaningful alignment. Since the IMGT numbering provides clear information about the position and length of these regions, it is very easy to omit them.
In the following screenshot, genes from the from the mouse IGHV1 subgroup have been selected :
At the bottom of this page,
- the "Add more sequences" button allows to select and add more sequences from other species, groups or subgroups. Note that this procedure can be repeated this as many times as needed.
- the "Calculate" button allows to calculate the synonymous non synonymous substitution rates between sequences (slow). At the top of the IMGT/PhyloGene substitution rates result page are shown the average values for the Ks and Ka of all the analysed sequences. The Ks and Ka between the sequences, compared two by two, are displayed in the table below.
- the "Download" button allows to download the alignment in the format of your choice (PHYLIP or FASTA).
When all the sequences to analyse have been included into the selection, press the "Compute a distance matrix" button to proceed to the next step.

IMGT/PhyloGene distance matrix result page

This page is only meant to show the distance matrix that has been calculated from the selected sequences.
Distance between DNA sequences belonging to the selection are calculated using the Kimura 2-parameters model of evolution [4]. Although rather simple, this model of evolution is very often used in phylogenetic analysis. For example, it comes as a default option in the DNADIST program from the PHYLIP package [6].
In the following screenshot, the distance matrix between genes from the mouse IGHV1 subgroups is shown :
Pressing the "Build a tree using NJ" button will proceed to the next step and build a phylogenetic tree from this distance matrix.

IMGT/PhyloGene tree drawing page

This page shows the phylogenetic tree that has been calculated using the Neighbor-Joining algorithm [5]. The NJ algorithm is one of the most famous algorithm for reconstructing phylogenetic trees. Besides giving good results in term of reconstruction, it is very fast, and therefore very well suited to online phylogenetic analysis, where nearly immediate response times are needed. Interested readers should refer to the literature ([5] [7] [8] for example) for further details about the NJ algorithm.
Once constructed, the NJ tree is rooted using the midpoint procedure, which consists in locating the root halfway in the tree from the most divergent sequences. Another way of rooting a phylogenetic tree is to include an outgroup in the form of one or several sequences whose duplication or speciation is anterior to every other duplication or speciation that has occurred among the rest of the selection. This can be done easily with the selection page described above (by pasting that (or those) sequence(s) in the text area, for example).
Once rooted, the tree is then drawn in the form of a horizontal phenogram. This kind of representation, very common in the literature, allows for easy visualisation of sequence groupings within the tree.
The default tree representation, IMGT/PhyloGene "tree with branch lengths", incorporates branch lengths : this means that the branch lengths of the represented tree have a meaning in term of evolutionary distance.
In the following screenshot, the NJ tree of the mouse IGHV1 subgroups is represented with branch lengths :
At the bottom of this page, a button allows you to draw the same "tree without branch lengths" : in that kind of representation, branch lengths are artificially arranged so as to align gene names and have no meaning in term of evolutionary distance.

IMGT/PhyloGene tree alignment page

In addition to the tree, it is possible to view, along with gene names, the sequences corresponding to the CDR1-IMGT, FR1-IMGT, CDR2-IMGT, FR2-IMGT, CDR3-IMGT, FR3-IMGT and V-RS (recombination signal including V-HEPTAMER, V-SPACER and V-NONAMER in 3' of V-REGION of a V-GENE) when available. Simply select the labels to be shown using the popup menu at the bottom of the page.
Note that the CDR-IMGT are omitted when the distance matrix is computed, they are still stored for later use. The same applies to the V-RS, which were stored at selection time. For this reason, they can now be added to the tree representation.
This kind of representation may yield interesting clues regarding the evolution of certain regions of the selected sequences. For example, in the following screenshot, where the CDR1-IMGT are represented in addition to the tree computed from the mouse IGHV1 sequences, the length of CDR1-IMGT nearly doubled (or was divided by two) in mouse sequences just after human-mouse speciation.

References

[1] Elemento O and Lefranc M-P. IMGT/PhyloGene, an online software package for phylogenetic analysis of immunoglobulin and T cell receptor genes. Dev. Comp. Immunol., 27, 763-779 (2003), PMID:12818634, LIGM:273.
[2] Elemento O. Phylogénie de familles multigéniques: Immunoglobulines et récepteurs T dans IMGT. Thèse de doctorat, Université Montpellier II, 2002.
[3] Lefranc M-P. IMGT, the international ImMunoGeneTics database®. Nucleic Acids Res., 31, 307-310 (2003), PMID:12520009, LIGM:268.
[4] Kimura M. A simple model for estimating evolutionary rates of base substitutions through comparative studies of nucleotide sequences. J. Mol. Evol., 16, 111-120 (1980), PMID:7463489.
[5] Saitou N and Nei M. The neighbor-joining method: a new method for reconstructing phylogenetic trees. Mol. Biol. Evol., 4, 1073-1095 (1987), PMID:3447015.
[6] Felsenstein J. PHYLIP - PHYLogeny Inference Package. Cladistics, 5, 164-166 (1989).
[7] Li W. Molecular Evolution. Sinauer Inc. Sunderland, Massachussets (1997).
[8] Swofford D, Olsen P, Waddell P and Hillis D. Molecular Systematics. Chapter Phylogenetic Inference, Sinauer Inc. Sunderland, Massachussets, 407-514 (1996).

IMGT/PhyloGene implementation and developments

The IMGT/PhyloGene tool was implemented by Olivier Elemento (LIGM, Montpellier, France, Ph.D. thesis, Université Montpellier II 02/12/2002 [1,2])
using Perl and C programs.

Céline Protat added C-DOMAIN sequences into the IMGT/PhyloGene database and adapted the IMGT/PhyloGene tool to deal with them in July 2003.

The IMGT/Phylogene tool has been improved by François Ehrenmann in October 2006.

Created:: 10/02/2002
Last updated:: 01/10/2006
Authors:: Olivier Elemento, François Ehrenmann, Céline Protat and Marie-Paule Lefranc; Contact: Marie-Paule Lefranc Marie-Paule.Lefranc@igh.cnrs.fr
Editors:: Quentin Kaas (2001-2007), François Ehrenmann (2006-2011), Caroline Tournier (2011-2013), Typhaine Paysan-Lafosse (2013-2016) and Patrice Duroux (LIGM, Montpellier, France)