Rockefeller University

Program for the Human Environment

Indicator Vector and TreeParser Software for Generating Klee diagrams

Klee diagrams generated by indicator vector analysis are alternatives to trees for representing nucleotide sequence alignments (Sirovich et al 2009, 2010). Indicator vectors are digital representations of nucleotide sequences in vector space, and can represent single sequences or sets of sequences. Indicator vector correlations, which are similar but not identical to distances, are displayed in a Klee diagram, a heat map of the indicator vector correlation matrix. The indicator vector/Klee method is computationally and visually scalable to large sets of aligned sequences–it readily displays patterns among 10K sequences in a single-page figure, for example. 

What follows is a step-by-step protocol for applying indicator vector analysis to a set of aligned nucleotide sequences.

The order of sequences in an alignment does not affect indicator vector calculations. However, for the resulting Klee diagram, it is useful to arrange the sequences to approximate evolutionary relationships. Here we provide a web-based program, TreeParser, that arranges aligned sequences to follow the order of terminals in a phylogenetic tree. TreeParser input files are 1) an aligned fasta file and 2) a Newick tree file in text format, and the output is a re-ordered .fas file. The example below utilizes a neighbor-joining (NJ) tree generated in MEGA as the template for ordering sequences.

Sirovich L, Stoeckle MY, Zhang Y (2009) A scalable method for analysis and display of DNA sequences. PLoS ONE 4:e7051.
Sirovich L, Stoeckle MY, Zhang Y (2010) Structural analysis of biodiversity. PLoS ONE 5:e9266.

Step 1: Run FASTA file through MEGA to create a phylogenetic tree. Under File menu, select “Write Tree in Table Format”, then save output as text file. 

Step 2: Upload your FASTA file and the MEGA tree text file to TreeParser software using windows below. (Large files may take a few minutes to process)

FASTA file 

Tree file 

Position of unique sequence identifier in .fas file header:

 1    2    3    4    Use entire header of each FASTA sequence  

sample .fas file header:
>position1|position2|position3|position4|

OR

 Run sample data (ignore file uploads and parameter. View sample Fasta and sample Tree files)

Download results

Step 3: Upload TreeParser output FASTA file to the Indicator Vector program.

Downloads:

Download the source code for Tree Parser (PHP saved as .txt, change to .php extension) here

Download the source code for Indicator Vector Klee Diagram (Matlab) here

MEGA software can be downloaded here. Click on the images below for instructions on using MEGA.

Start MEGA and input Nucleotide Sequence data.
Select your analysis preferences. No Test of Phylogeny, A Kimura 2-parameter model, and Pairwise Deletion are recommended.
Generate a phylogenetic tree.
Export the tree as a text file. This file will be used as input in the Tree Parser program.

About the Bar Code of Life site
This web site is an outgrowth of the Taxonomy, DNA, and Barcode of Life meeting held at Banbury Center, Cold Spring Harbor Laboratory, September 9-12, 2003. It is managed by Mark Stoeckle at the Program for the Human Environment (PHE) at The Rockefeller University.

Contact: mark.stoeckle@rockefeller.edu

About the Program for the Human Environment
The involvement of the Program for the Human Environment in DNA barcoding dates to Jesse Ausubel's attendance in February 2002 at a conference in Nova Scotia organized by the Canadian Center for Marine Biodiversity. At the conference, Paul Hebert presented for the first time his concept of large-scale DNA barcoding for species identification. Impressed by the potential for this technology to address difficult challenges in the Census of Marine Life, Jesse agreed with Paul on encouraging a conference to explore the contribution taxonomy and DNA could make to the Census as well as other large-scale terrestrial efforts. In his capacity as a Program Director of the Sloan Foundation, Jesse turned to the Banbury Conference Center of Cold Spring Harbor Laboratory, whose leader Jan Witkowski prepared a strong proposal to explore both the scientific reliability of barcoding and the processes that might bring it to broad application. Concurrently, PHE researcher Mark Stoeckle began to work with the Hebert lab on analytic studies of barcoding in birds. Our involvement in barcoding now takes 3 forms: assisting the organizational development of the Consortium for the Barcode of Life and the Barcode of Life Initiative; contributing to the scientific development of the field, especially by studies in birds, and contributing to public understanding of the science and technology of barcoding and its applications through improved visualization techniques and preparation of brochures and other broadly accessible means, including this website. While the Sloan Foundation continues to support CBOL through a grant to the Smithsonian Institution, it does not provide financial support for barcoding research itself or support to the PHE for its research in this field.