Transcript Structural Bioinformatics

Forces and Prediction of Protein
Structure
Ming-Jing Hwang (黃明經)
Institute of Biomedical Sciences
Academia Sinica
http://gln.ibms.sinica.edu.tw/
Science 2005
Sequence - Structure - Function
MADWVTGKVTKVQ
NWTDALFSLTVHAP
VLPFTAGQFTKLGLE
IDGERVQRAYSYVN
SPDNPDLEFYLVTVP
DGKLSPRLAALKPG
DEVQVVSEAAGFFV
LDEVPHCETLWMLA
TGTAIGPYLSILR
Sequence/Structure Gap
Current (May 15, 2007) entries in protein sequence and structure
database:
SWISS-PROT/TREMBL : 267,354/4,361,897
PDB : 43,459
Sequence
Number of entries


Structure
Year
Structural Bioinformatics:
Sequence/Structure Relationship
Percent Identity
100
90
All possible sequences of amino acids
80
Protein structures
observed in nature
70
60
50
40
30
20
Protein sequences
observed in nature
Twilight zone
Midnight zone 10
0
Structure Prediction Methods
Homology modeling
Fold recognition
ab initio
0
10
20
30
40
50
60
70
80
90 100
% sequence identity
Levinthal’s paradox (1969)


If we assume three possible states for every flexible
dihedral angle in the backbone of a 100-residue protein,
the number of possible backbone configurations is 3200.
Even an incredibly fast computational or physical
sampling in 10-15 s would mean that a complete sampling
would take 1080 s, which exceeds the age of the universe
by more than 60 orders of magnitude.
Yet proteins fold in seconds or less!
Berendsen
Energy landscapes of protein folding
Borman, C&E News, 1998
Levitt’s lecture for S*
Levitt
Levitt
Other factors






Formation of 2nd elements
Packing of 2nd elements
Topologies of fold
Metal/co-factor binding
Disulfide bond
…
Ab initio/new fold prediction
Physics-based (laws of physics)
 Knowledge-based (rules of evolution)

Levitt
Levitt
Levitt
Levitt
Levitt
Levitt
Levitt
Levitt
Levitt
Levitt
Levitt
Levitt
Levitt
Molecular Mechanics (Force Field)
Levitt
1-microsecond 980ns
MD simulation
- villin headpiece
- 36 a.a.
- 3000 H2O
- 12,000 atoms
- 256 CPUs (CRAY)
-~4 months
- single trajectory
Duan & Kollman, 1998
Protein folding by MD
PROTEIN FOLDING:
A Glimpse of the Holy Grail?
Herman J. C. Berendsen*
"The Grail had many different manifestations
throughout its long history, and many have
claimed to possess it or its like". We might have
seen a glimpse of it, but the brave knights must
prepare for a long pursuit.
Massively distributed computing
SETI@home:
 Folding@home
 Distributed folding
 Sengent’s drug design
 FightAIDS@home
…

Massively distributed computing
Letters to nature (2002)
- engineered protein (BBA5)
- zinc finger fold (w/o metal)
- 23 a.a.
- solvation model
- thousands of trajectories each
of 5-20 ns, totaling 700 ms
- Folding@home
- 30,000 internet volunteers
- several months, or ~a million
CPU days of simulation
Energy landscapes of protein folding
Borman, C&E News, 1998
Protein-folding prediction technique
CGU: Convex Global
Underestimation
- K. Dill’s group
Challenges of physics-based methods
Simulation time scale
 Computing power
 Sampling
 Accuracy of energy functions

Structure Prediction Methods
Homology modeling
Fold recognition
ab initio
0
10
20
30
40
50
60
70
80
90 100
% sequence identity
Flowchart of homology (comparative) modeling
From Marti-Renom et al.
Fold recognition
Find, from a library of folds, the 3D template
that accommodates the target sequence best.
Also known as “threading” or “inverse folding”
Useful for twilight-zone sequences
Fold recognition (aligning sequence to
structure)
(David Shortle, 2000)
3D->1D score
On X-ray, NMR, and computed models
(Rost, 1996)
Reliability and uses of comparative models
Marti-Renom et al. (2000)
Pitfalls of comparative modeling



Cannot correct alignment errors
More similar to template than to true
structure
Cannot predict novel folds
Ab initio/new fold prediction
Physics-based (laws of physics)
 Knowledge-based (rules of evolution)

From 1D  2D  3D
Primary
LGINCRGSSQCGLSGGNLMVRIRDQACGNQGQTWCPGERRAKVCGTGNSISAY
VQSTNNCISGTEACRHLTNLVNHGCRVCGSDPLYAGNDVSRGQLTVNYVNSC
seq. to str. mapping
Secondary
(fragment)
Tertiary
fragment assembly
CASP Experiments
One group dominates the ab initio
(knowledge-based) prediction
One lab dominated in CASP4
Some CASP4 successes
Baker’s group
Ab initio structure prediction server
Toward High-Resolution de Novo Structure
Prediction for Small Proteins
--Philip Bradley, Kira M. S. Misura, David Baker
(Science 2005)
The prediction of protein structure from
amino acid sequence is a grand challenge of
computational molecular biology. By using a
combination of improved low- and highresolution conformational sampling methods,
improved atomically detailed potential
functions that capture the jigsaw puzzle–like
packing of protein cores, and highperformance computing, high-resolution
structure prediction (<1.5 angstroms) can be
achieved for small protein domains (<85
residues). The primary bottleneck to
consistent high-resolution prediction appears
to be conformational sampling.
3D to 1D?
Science 2003
A computer-designed protein (93 aa)
with 1.2 A resolution
Structure prediction servers
http://bioinfo.pl/cafasp/list.html
Hybrid approach for solving macromolecular
complex structures
Thank You!