Description

This track shows multiple alignments of 8 vertebrate species with a measure of evolutionary conservation, based on a phylogenetic hidden Markov model, phastCons (Siepel et al., 2005).

The multiple alignments were generated using multiz and other tools in the UCSC/Penn State Bioinformatics comparative genomics alignment pipeline. The conservation measurements were created using the phastCons package from Adam Siepel at Cornell University.

Multiz alignments of the following assemblies were used to generate this track:

OrganismSpeciesRelease dateUCSC versionalignment type
OrangutanPongo pygmaeus abelii July 2007ponAbe2reference species
HumanHomo sapiens Mar 2006 hg18syntenic net
ChimpanzeePan troglodytes Mar 2006 panTro2syntenic net
RhesusMacaca mulatta Jan 2006 rheMac2syntenic net
MarmosetCallithrix jacchusJune 2007 calJac1reciprocal best net
MouseMus musculus Jul 2007 mm9syntenic net
OpossumMonodelphis domestica Jan 2006 monDom4chain net
PlatypusOrnithorhychus anatinus Mar 2007 ornAna1chain net

Table 1. Genome assemblies included in the 8-way Conservation track.

Display Conventions and Configuration

In full and pack display modes, conservation scores are displayed as a wiggle track (histogram) in which the height reflects the size of the score. The conservation wiggle can be configured in a variety of ways to highlight different aspects of the displayed information. Click the Graph configuration help link for an explanation of the configuration options.

Pairwise alignments of each species to the $organism genome are displayed below the conservation histogram as a grayscale density plot (in pack mode) or as a wiggle (in full mode) that indicates alignment quality. In dense display mode, conservation is shown in grayscale using darker values to indicate higher levels of overall conservation as scored by phastCons.

Checkboxes on the track configuration page allow selection of the species to include in the pairwise display. Note that excluding species from the pairwise display does not alter the the conservation score display.

To view detailed information about the alignments at a specific position, zoom the display in to 30,000 or fewer bases, then click on the alignment.

Gap Annotation

The Display chains between alignments configuration option enables display of gaps between alignment blocks in the pairwise alignments in a manner similar to the Chain track display. The following conventions are used:

Genomic Breaks

Discontinuities in the genomic context (chromosome, scaffold or region) of the aligned DNA in the aligning species are shown as follows:

Base Level

When zoomed-in to the base-level display, the track shows the base composition of each alignment, with bases in annotated repetitive elements displayed in lower case. The numbers and symbols on the Gaps line indicate the lengths of gaps in the $organism sequence at those alignment positions relative to the longest non-$organism sequence. If there is sufficient space in the display, the size of the gap is shown. If the space is insufficient and the gap size is a multiple of 3, a "*" is displayed; other gap sizes are indicated by "+".

Codon translation is available in base-level display mode if the displayed region is identified as a coding segment. To display this annotation, select the species for translation from the pull-down menu in the Codon Translation configuration section at the top of the page. Then, select one of the following modes:

Codon translation uses the following gene tracks as the basis for translation, depending on the species chosen (Table 2). Species listed in the row labeled "None" do not have species-specific reading frames for gene translation.

Gene TrackSpecies
UCSC Geneshuman, mouse
Ensembl Geneschimp,rhesus, opossum, platypus
Noneorangutan, marmoset
Table 2. Gene tracks used for codon translation.

Methods

Pairwise alignments with the $organism genome were generated for each species using blastz from repeat-masked genomic sequence. Lineage-specific repeats were removed prior to alignment, then reinserted. Pairwise alignments were then linked into chains using a dynamic programming algorithm that finds maximally scoring chains of gapless subsections of the alignments organized in a kd-tree. The scoring matrix and parameters for pairwise alignment and chaining were tuned for each species based on phylogenetic distance from the reference. High-scoring chains were then placed along the genome, with gaps filled by lower-scoring chains, to produce an alignment net. For more information about the chaining and netting process and parameters for each species, see the description pages for the Chain and Net tracks.

An additional filtering step was introduced in the generation of the 8-way conservation track to reduce the number of paralogs and pseudogenes from the high-quality assemblies and the suspect alignments from the low-quality assemblies: the pairwise alignments of high-quality mammalian sequences (placental and marsupial) were filtered based on synteny; those for 2X mammalian genomes were filtered to retain only alignments of best quality in both the target and query ("reciprocal best").

The resulting best-in-genome pairwise alignments were progressively aligned using multiz/autoMZ, following the tree topology diagrammed above, to produce multiple alignments. The multiple alignments were post-processed to add annotations indicating alignment gaps, genomic breaks, and base quality of the component sequences. The annotated multiple alignments, in MAF format, are available for bulk download. An alignment summary table containing an entry for each alignment block in each species was generated to improve track display performance at large scales. Framing tables were constructed to enable visualization of codons in the multiple alignment display.

Conservation scoring was performed using the PhastCons package (A. Siepel), which computes conservation based on a two-state phylogenetic hidden Markov model (HMM). PhastCons measurements rely on a tree model containing the tree topology, branch lengths representing evolutionary distance at neutrally evolving sites, the background distribution of nucleotides, and a substitution rate matrix. The <<<<<<< multiz8way.html >>>>>> 1.5 TARGET=_blank>vertebrate tree model for this track was generated using the phyloFit program from the phastCons package (REV model, EM algorithm, medium precision) using multiple alignments of 4-fold degenerate sites extracted from the 28-way human(hg18) alignment (msa_view). The 4d sites were derived from the Oct 2005 Gencode Reference Gene set, which was filtered to select single-coverage long transcripts. The phastCons parameters used for the conservation measurement were: expected-length=45, target-coverage=.3 and rho=.31

The phastCons program computes conservation scores based on a phylo-HMM, a type of probabilistic model that describes both the process of DNA substitution at each site in a genome and the way this process changes from one site to the next (Felsenstein and Churchill 1996, Yang 1995, Siepel and Haussler 2005). PhastCons uses a two-state phylo-HMM, with a state for conserved regions and a state for non-conserved regions. The value plotted at each site is the posterior probability that the corresponding alignment column was "generated" by the conserved state of the phylo-HMM. These scores reflect the phylogeny (including branch lengths) of the species in question, a continuous-time Markov model of the nucleotide substitution process, and a tendency for conservation levels to be autocorrelated along the genome (i.e., to be similar at adjacent sites). The general reversible (REV) substitution model was used. Unlike many conservation-scoring programs, note that phastCons does not rely on a sliding window of fixed size; therefore, short highly-conserved regions and long moderately conserved regions can both obtain high scores. More information about phastCons can be found in Siepel et al. 2005.

PhastCons currently treats alignment gaps as missing data, which sometimes has the effect of producing undesirably high conservation scores in gappy regions of the alignment. We are looking at several possible ways of improving the handling of alignment gaps.

Credits

This track was created using the following programs:

The phylogenetic tree is based on Murphy et al. (2001) and general consensus in the vertebrate phylogeny community as of March 2007.

References

Phylo-HMMs and phastCons:

Felsenstein J, Churchill GA. A Hidden Markov Model approach to variation among sites in rate of evolution. Mol Biol Evol. 1996 Jan;13(1):93-104.

Siepel A, Haussler D. Phylogenetic Hidden Markov Models. In R. Nielsen, ed., Statistical Methods in Molecular Evolution, pp. 325-351, Springer, New York (2005).

Siepel A, Bejerano G, Pedersen JS, Hinrichs AS, Hou M, Rosenbloom K, Clawson H, Spieth J, Hillier LW, Richards S, et al. Evolutionarily conserved elements in vertebrate, insect, worm, and yeast genomes. Genome Res. 2005 Aug;15(8):1034-50.

Yang Z. A space-time process model for the evolution of DNA sequences. Genetics. 1995 Feb;139(2):993-1005.

Chain/Net:

Kent WJ, Baertsch R, Hinrichs A, Miller W, Haussler D. Evolution's cauldron: duplication, deletion, and rearrangement in the mouse and human genomes. Proc Natl Acad Sci U S A. 2003 Sep 30;100(20):11484-9.

Multiz:

Blanchette M, Kent WJ, Riemer C, Elnitski L, Smit AF, Roskin KM, Baertsch R, Rosenbloom K, Clawson H, Green ED, et al. Aligning multiple genomic sequences with the threaded blockset aligner. Genome Res. 2004 Apr;14(4):708-15.

Blastz:

Chiaromonte F, Yap VB, Miller W. Scoring pairwise genomic sequence alignments. Pac Symp Biocomput. 2002;:115-26.

Schwartz S, Kent WJ, Smit A, Zhang Z, Baertsch R, Hardison RC, Haussler D, Miller W. Human-mouse alignments with BLASTZ. Genome Res. 2003 Jan;13(1):103-7.

Phylogenetic Tree:

Murphy WJ, Eizirik E, O'Brien SJ, Madsen O, Scally M, Douady CJ, Teeling E, Ryder OA, Stanhope MJ, de Jong WW, Springer MS. Resolution of the early placental mammal radiation using Bayesian phylogenetics. Science. 2001 Dec 14;294(5550):2348-51.