Molecular dynamics of two ester
hydrolases: comparison using the
BioSimGrid database
Kaihsu Tai
The systems
● both serine esterases (ester hydrolases, EC 3.1.1)
● with the catalytic triad: Ser - His - Glu|Asn
OMPLA AChE
N156
H142
S144
E334
H447
S203
The trajectoriestrajectory description time/ns id code
Apo-AChE unliganded monomer 15 GB-OXF_11
Fas-AChE liganded with fasciculin 2 5 GB-OXF_12
OMPLA1 unliganded monomer, Ewald 10 GB-OXF_16
OMPLA2 unliganded dimer, Ewald 10 GB-OXF_17
OMPLA3 liganded dimer, Ewald 10 GB-OXF_15
OMPLA4 unliganded monomer, cutoff 5.7 GB-OXF_9
OMPLA5 unliganded dimer, cutoff 5 GB-OXF_10
OMPLA6 liganded dimer, cutoff 5 GB-OXF_14
● ...not yet data deluge
● but...
articles with “molecular dynamics” in
title or abstract in the Biophysical
Journal
(one of the first protein MD: 1980 McCammon et al.)
year number
2004 (up to November) ~97
2003 ~96
2002 ~70
2001 ~55
2000 ~471995 to 1999 ~186
1990 to 1994 ~48
1976 to 1989 <25
Things we tried that did not work
very well● Picking candidate atoms from the side chain of
the triad and generate
– the meta-plot
– the meta-histogram
Meta-plot
Meta-histogram
A set of metrics for triad intactness
A set of metrics for triad intactness
Looking ahead
● Here we had some intuition, and formed a model.
– Is it possible to do model-free data mining?
– n × n rmsd (as Bing may have shown), etc.
– Yes, we can do it, but what is the
biological/biochemical relevance?
● The answer is 42 - what is the question?