Jim Garey (garey at next.duq.edu) wrote:
:: I need some information on the mysterious "taking secondary structure
: into account" when creating multiple alignments of 18S rRNA genes.
: My educated guess on the process is to mark known stem sequences on
: one sequence in the alignment, ie human, and then look for
: palindromic sequences in similar regions of the other sequences in
: the alignment and adjust the alignment to match them with the known
: stem. Is this how it is done?
In an alignment both primary and secondary structure should be similar
in a certain region. It is easy to align very conserved areas on basis
of primary structure alone. In the more variable areas the alignment is
often not that clear. Here you can use the secondary structure information
to improve the alignment in a sort of iterative way. You can align on basis
of primary structure, and check whether the general sec. str. can be
transposed. If it cannot, you can alter the alignment a bit. Or when the
alignment on one helix segment is correct, you can search the complement.
We do this study using DCSE, which is a multiple alignment editor which
lets you incorporate the structure information in the alignment by insertion
of special symbols. These structures can be shown in color as well.
[Plug: DCSE is available on host uiam3.uia.ac.be under the directory dcse
do 'cd [.dcse] to get there. The old version is available for DOS, VMS, and
SGI. A new version for X-Windows is available for SGI.]
: does anyone have any good references for this?
:: Can anyone direct me to references containing good information on
: known 18S secondary structures (I am interested in yeast and
:: Thanks in advance for the help. Jim Garey garey at next.duq.edu
Peter De Rijk