Database searching

1
Data-base searching Multiple alignment Phylogene
tic Reconstruction Protein structure prediction
McClure, 2006
2
Multiple protein sequence alignment uses in
structure determination
McClure, 2006
3
If the protein of interest has a relative whose
structure has been determined, then structure
prediction is much easier.
McClure, 2006
4
Solved structures and their relatives
http//www.rcsb.org/pdb/home/home.do http//scop.
mrc-lmb.cam.ac.uk/scop/ http//www.cathdb.info/la
test/index.html
5

• Predicting protein structure
• a) primary
• b) secondary
• c) tertiary
• d) quaternary
• Predicting protein folds
• Predicting protein disorder
• Predicting protein interactions

McClure, 2006
6
What do we already know about proteins?
The three-dimensional shape or
conformation
is stabilized by non-covalent forces or
hydrogen bonds
denatured
native
PRIMARY STRUCTURE
McClure, 2006
7
McClure, 2006
8
(No Transcript)
9
Beta-sheets can take on a higher order
structures saddles or barrels depending on which
way they turn
10
Predicting 3-dimensional structure de novo is
very difficult. So the best methods use a
prior information For homology and
threading approaches.
McClure, 2006
11
There are two basic tertiary shapes that the
secondary structure adopts.
GLOBS
FIBERS
http//www.uccs.edu/rmelamed/MicroFall2002/Chapte
r202/protein20types.jpg
McClure, 2006
12
The tertiary conformation is the shape the
secondary structure regions adopt alpha
helices, beta-pleated sheets and regions of
disorder.
McClure, 2006
13
Like secondary-structure the tertiary
conformation is stabilized mostly by
non-covalent bonds but some covalent bonds are
integral to the higher-order structure. Non-coval
ent Covlaent Eletrostatic interations
Disulphide bridges Hydrogen bonds Isopeptide
linkage Hydrophobic interations
McClure, 2006
14
Natively unfolded proteins have been shown to
have little or no secondary structure until
contact is made with the physiological partner
and the same has been demonstrated for regions
of disorder or random coil within a given
protein.
McClure, 2006
15
(No Transcript)
16

• Assumptions of Homology Modeling
• Structure is more conserved than sequence.
• Amino acid replacements are compensatory.
• 3) Secondary structure is conserved.
• Sequence identity considerations
• 1) If sequence identity is gt70 to sequence of
  known
• structure than a high resolution model lt3A can
  be built.
• 2) A query sequence must have gt25 identity with
• a known structure for classical methods.
• Prediction Limitations
• 1) Loops are not highly-conserved in both
  sequence or length
• 2) Surface side chains highly variable

17
(No Transcript)
18
(No Transcript)
19
When sequences are lt25 ID to a
structure or there is no structure 3D prediction
is by threading also called fold mapping
or distant homology modeling
20
(No Transcript)
21
(No Transcript)
22
Online Resources Baylor College of Medicine
Launcher http//searchlauncher.bcm.tmc.edu/seq-se
arch/struc-predict.html Structure Prediction
Online Resources SPORes http//cgat.ukm.my/spores
/
DISOPRED server http//bioinf.cs.ucl.ac.uk/disopr
ed/
Lists current links for all kinds of online
protein prediction sites http//www.up.univ-mrs.f
r/wabim/english/logligne.htmlpredi
McClure, 2006