This track shows alignments of the human genome with itself, using
a gap scoring system that allows longer gaps than traditional
affine gap scoring systems. The system can also tolerate gaps
in both sets of sequence simultaneously. After filtering out the
"trivial" alignments produced when identical locations of the
genome map to one another (e.g. chrN mapping to chrN),
the remaining alignments point out areas of duplication within the
human genome. The pseudoautosomal regions of chrX and chrY are an
exception: in this assembly, these regions have been copied from chrX into
chrY, resulting in a large amount of self chains aligning in these positions
on both chromosomes.
The chain track displays boxes joined together by either single or
double lines. The boxes represent aligning regions. Single lines indicate
gaps that are largely due to a deletion in the query assembly or an
insertion in the target assembly. Double lines represent more complex gaps
that involve substantial sequence in both the query and target assemblies.
This may result from inversions, overlapping deletions, an abundance of local
mutation, or an unsequenced gap in one of the assemblies. In cases where
multiple chains align over a particular region of the human genome, the
chains with single-lined gaps are often due to processed pseudogenes, while
chains with double-lined gaps are more often due to paralogs and unprocessed
Chains have both a score and a normalized score. The score is derived by
comparing sequence similarity, while penalizing both mismatches and gaps
in a per base fashion. This leads to longer chains having greater scores,
even if a smaller chain provides a better match. The normalized score divides
the score by the length of the alignment, providing a more comparable score value
not dependent on the match length.
Display Conventions and Configuration
By default, the chains are colored by the normalized score. This can be changed
to color based on which chromosome they map to in the aligning organism. There is also
an option to color all the chains black.
To display only the chains of one chromosome in the aligning
organism, enter the name of that chromosome (e.g. chr4) in box next to:
Filter by chromosome.
By default, chains with a score of 20,000 or more are displayed. This default value provides
a conservative cutoff, filtering out many false-positive alignments with low sequence
similarity, or high penalties. It should be noted however, that alignments below this
threshold may still be indicative of homology.
In the "pack" and "full" display
modes, the individual feature names indicate the chromosome, strand, and
location (in thousands) of the match for each matching alignment.
The genome was aligned to itself using blastz. Trivial alignments were
filtered out, and the remaining alignments were converted into axt format
using the lavToAxt program. The axt alignments were fed into axtChain, which
organizes all alignments between a single target chromosome and a single
query chromosome into a group and creates a kd-tree out of the gapless
subsections (blocks) of the alignments. A dynamic program was then run over
the kd-trees to find the maximally scoring chains of these blocks. Chains
scoring below a threshold were discarded; the remaining chains are displayed
in this track.
Blastz was developed at Pennsylvania State University by
Minmei Hou, Scott Schwartz, Zheng Zhang, and Webb Miller with advice from
Lineage-specific repeats were identified by Arian Smit and his
The axtChain program was developed at the University of California
at Santa Cruz by Jim Kent with advice from Webb Miller and David Haussler.
The browser display and database storage of the chains were generated
by Robert Baertsch and Jim Kent.
Chiaromonte F, Yap VB, Miller W.
Scoring pairwise genomic sequence alignments.
Pac Symp Biocomput 2002, 115-26 (2002).
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.
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.