is a novel parameter-free method for de novo computational
motif discovery, designed for biologists. The only inputs required by
SCOPE are a set of unaligned sequences and a species designation. SCOPE utilizes three algorithms to identify
sequence motifs: BEAM finds non-degenerate
motifs, PRISM finds degenerate motifs, and SPACER finds bipartite
Results from all three algorithms are merged and the best scoring
motifs are presented.
tests on 78 sets of coregulated genes with experimentally determined
binding sites, SCOPE identified motifs (of previously unknown length,
degree of degeneracy and orientation) with a high level of accuracy.
Ten other motif-finding programs with web-based interfaces were also
run on this dataset for comparison, and SCOPE was found to be more
accurate than all of them. SCOPE is also extremely robust to noise in
input datasets (for details see our SCOPE paper: Chakravarty et al., 2006 submitted). In
semi-synthetic test regulons where four times as many random sequences
as bona fide upstream sequences were present, SCOPE's performance
degraded by only 21%, thus SCOPE is able to predict transcription
factor binding sites with a high level of accuracy.
accepts genes for analysis either as FASTA sequences or as gene names,
and is currently configured for the analysis of a number of
prokaryotic and eukaryotic genomes. Run times for SCOPE are fast (in
the order of one to five minutes), and output consists of the
significance score, the location, count, and identity (consensus
sequence and PWM) of all statistically significant motifs. Users may
modify the output of SCOPE to individually view the most significant
motifs from each of its component algorithms and view maps of the
locations of each of the motifs in the set of genes examined.
the SCOPE web server is a practical resource for members of the
biological community with an interest in computational motif-finding.
home page: http://genie.dartmouth.edu/scope
system(s): Platform independent
requirements: Java 1.5 or higher
- License: Free for academic use,
- Sourcecode, JavaDoc and executables
are available here.