Shoba Ranganathan, Ph.D.
Associate Professor
Dept. of Biochemistry,
Dept. of Biological Sciences &
Bioinformatics Centre Tel:
+65-6874-3566
National University of Singapore +65-6774-7149
10 Kent Ridge Crescent
Fax: +65-6778-2466
Singapore 119260 email:
shoba@bic.nus.edu.sg
Aligning mRNA/EST data to geneomic sequences - a novel approach
Alternative splicing is a biological phenomenon that produces several
different transcripts from a single gene, thereby having the potential
to
produce several protein products from one gene. To study this phenomenon,
information on the intron-exon arrangement of a gene is essential,
usually
obtained by aligning mRNA/EST sequences to their cognate genomic
sequences.
We will present a novel rapid approach, MGAlign, for the alignment
of an
mRNA/EST sequence to its genomic sequence. Current alignment techniques
require the entire genomic sequence to be searched repetitively
for high
scoring local alignments. Our approach uses a known biological
characteristic of this particular type of alignment problem to speed
up the
alignment, namely that the region of genomic sequence occupied by
an
mRNA/EST sequence is much shorter than the total length of the genomic
sequence. By rapidly converging on the region where mRNA/EST alignments
are
to be expected in the genomic sequence, we have achieved speed enhancements
of several folds over comparable methods available currently. The
program
entitled MGAlign (implemented in C++) was tested and shown to be
much faster
and more accurate than current alignment methods that search the
entire
genomic sequence. The low memory requirement of this
approach also allows
implementation of this program for large genomic sequences even
on low-end
personal computers. A web server implementation as well as the binaries
for
the program are freely available at http://origin.bic.nus.edu.sg/mgalign.
We also provide a freely available web service, MGAlignIt
(http://origin.bic.nus.edu.sg/mgalign/mgalignit),
based on MGAlign. Besides
the alignment itself, this web service allows users to effectively
visualize
the alignment in a graphical manner and to perform limited analysis
on the
alignment output. The server also permits the alignment to be saved
in
several forms, both graphical and text, suitable for further processing
and
analysis by other programs.
|