BEGIN:VCALENDAR
VERSION:2.0
PRODID:-//Talks.cam//talks.cam.ac.uk//
X-WR-CALNAME:Talks.cam
BEGIN:VEVENT
SUMMARY:Some interesting data: Cryptic DNA sequence periodicities are ubiq
 uitous\, organism specific and distinguish introns from exons. - Giselle W
 alker
DTSTART:20070228T140000Z
DTEND:20070228T150000Z
UID:TALK6756@talks.cam.ac.uk
CONTACT:David MacKay
DESCRIPTION:Nucleotide bases in DNA are usually assumed to evolve independ
 ently.  However\, within a genome each of the 16 dinucleotide pairs tends 
 to be consistently either commoner or rarer than expected\, to the point w
 here together they constitute taxon-specific genomic signatures.  Here we 
 extend these observations to explore gapped dinucleotide motifs (GDMs): di
 nucleotide pairs with 0-9 intervening bases.  As with simple dinucleotides
 \, GDMs define signatures that are highly consistent across a genome. Tree
 s constructed from matrices of GDM similarity suggest that phylogenetic si
 gnal exists but can sometimes be swamped by convergence.  To find the like
 ly basis of GDM signatures we used a principal component analysis to summa
 rise GDM profiles and show that the dominant predictors of both the streng
 th and nature of the patterning are the %AT composition of the genome and 
 the optimal growth temperature of the organism\, particularly among prokar
 yotes.  For eukaryotes\, other factors such as genome size and chromosome 
 number exert a minor but significant influence.  Dividing eukaryotic seque
 nces into intronic and exonic sequences reveals a striking difference in G
 DM patterns\, exons being dominated by 3-base periodicities that are large
 ly absent from introns\, and from any of the whole genome sequence profile
 s.  By implication\, GDMs have potential to be used to identify anonymous 
 DNA fragments\, either to taxon among species\, or perhaps even by functio
 n within a species.  GDMs also illuminate constraints leading to convergen
 t evolution.  Either way\, the strength of the GDM profiles suggest that D
 NA evolution is less random than is often assumed.
LOCATION:TCM Seminar Room\, Cavendish Laboratory\, Department of Physics
END:VEVENT
END:VCALENDAR
