Multiple whole genome alignments and novel biomedical applications at the VISTA portal

Abstract

The VISTA portal for comparative genomics is designed to give biomedical scientists a unified set of tools to lead them from the raw DNA sequences through the alignment and annotation to the visualization of the results. The VISTA portal also hosts the alignments of a number of genomes computed by our group, allowing users to study the regions of their interest without having to manually download the individual sequences. Here we describe various algorithmic and functional improvements implemented in the VISTA portal over the last 2 years. The VISTA Portal is accessible at http://genome.lbl.gov/vista.

Figure 1.

SuperMap Algorithm. (a) Local alignment hits: regions A and B correspond to duplications in Organism 1; regions C and D correspond to duplications in Organism 2; (b) S-Lagan chain for Organism 1 as a base. Chain increases in direction of X-axis, but can jump up and down in Y-direction (Organism 2), region D is left out; (c) S-Lagan chain for Organism 2 as a base—chain increases in direction of Y-axis, region B is left out; (d) SuperMap output—combines regions of Figures b and c.

Figure 2.

Multiple alignment with LAGAN in the VISTA Genome Pipeline (VGP). After running a local alignment program, SuperMap Chaining is used to identify all rearrangements. The resulting regions are aligned with LAGAN, and finally a maximum matching algorithm is used to predict ancestral contigs. These ancestral contigs are then used to align to outgroup genomes is the higher levels of the phylogenetic tree.

Figure 3.

VISTA Browser display of 9.6 kb fragment of NR1H3 gene on Chr. 11 of the human genome (hg17). VISTA plots for the five-way Human–Dog–Mouse–Rat–Chicken alignment are shown. Conserved sequences in VISTA (70%/100 bp cutoff) are colored according to the annotation (exons—dark blue, UTRs—turquoise and non-coding—pink). Rank-VISTA peaks identified by Gumby (P < 0.5) are shown as vertical bars following the same coloring convention. At the bottom of the window one can see the base-pair browser and SNP data [dbSNP annotation and PolyPhen (5) prediction of functionality for coding SNPs].