Analysis of multiple genomic sequence alignments: a web resource, online tools, and lessons learned from analysis of mammalian SCL loci

Abstract

Comparative analysis of genomic sequences is becoming a standard technique for studying gene regulation. However, only a limited number of tools are currently available for the analysis of multiple genomic sequences. An extensive data set for the testing and training of such tools is provided by the SCL gene locus. Here we have expanded the data set to eight vertebrate species by sequencing the dog SCL locus and by annotating the dog and rat SCL loci. To provide a resource for the bioinformatics community, all SCL sequences and functional annotations, comprising a collation of the extensive experimental evidence pertaining to SCL regulation, have been made available via a Web server. A Web interface to new tools specifically designed for the display and analysis of multiple sequence alignments was also implemented. The unique SCL data set and new sequence comparison tools allowed us to perform a rigorous examination of the true benefits of multiple sequence comparisons. We demonstrate that multiple sequence alignments are, overall, superior to pairwise alignments for identification of mammalian regulatory regions. In the search for individual transcription factor binding sites, multiple alignments markedly increase the signal-to-noise ratio compared to pairwise alignments.

Figure 1

A long-range four-way mammalian alignment is capable of detecting all known SCL regulatory regions. A SynPlot representation of a human/dog/mouse/rat alignment of the SCL locus. Numbers on the horizontal axis do not refer to any one of the sequences in the alignment, but instead represent distance in nucleotides from the beginning of the aligned file (i.e., gaps in the alignment are counted). Numbers on the vertical axis represent an alignment score in a 100-bp window moved by 25-bp increments across the entire alignment. The horizontal black lines above the profile represent the human (H), dog (D), mouse (M), and rat (R) sequences and illustrate the position of gaps introduced to permit optimum alignment. Coding and noncoding exons are demonstrated by red and brown boxes, respectively, and repeats are illustrated by the smaller blue boxes. Regions of open chromatin and the +23 kb enhancer are shaded in green over the plot. Peaks of conserved sequence can be seen within each known regulatory region. Three well conserved peaks in regions of unknown function (-10 kb, +26 kb, and +30 kb) are also indicated. Pr1a, promoter 1a; Pr1b, promoter 1b.