Transcript HPC in Drug Discovery

Dr. Olivier Schwartz / Science Photo Library
Manuel C. Peitsch
Novartis
GRIDs in Drug Discovery
and Knowledge Management
The Challenges in Drug Discovery
“Drug Discovery suffers from a high attrition rate as many candidates
prove ineffective or toxic in the clinic, owing to a poor understanding
of the diseases, and thus the biological systems, they target”
}
Mechanism-based Drug Discovery
 Understanding Disease
 Pathways elucidation
 Target validation
Systems Biology:
Combination of *Omics &
Mathematical
Modelling
 Clinical PoC
New candidate drug with maximised therapeutic
window.
EGEE‘06 / M. Peitsch / Sept, 2006
Fully Leverage in silico sciences
Bioinformatics Lab
Target
finding
Text Informatics
Macromolecular
Structure & Function Lab
Target
validation
Lead
finding
Protein Modeling
Computational
Chemistry Lab
Lead
optim.
In silico Profiling
In Silico Drug Discovery Pipeline
Comparative Genomics
EGEE‘06 / M. Peitsch / Sept, 2006
HT Docking
In silico Combichem
In Silico Drug Discovery Pipeline: Can it be done?
1990
1995
Productive
Automated Protein
modelling email server
2000
2005
First PC-GRID
at Novartis
GeneCrunch
SETI@Home
Productive
Automated Protein
modelling Web server
3D-Crunch
Genome scale Automated
Protein modelling
Docking
in production
at Novartis
First
semi-automated
In Silico Drug
Discovery
Pipeline ?
Full Transcriptome
Modelling at Novartis
Protein Model
Structure database
SETI@Home recognised as a leading new concept (ComputerWorld Award)
GeneCrunch recognised as a leading new concept (ComputerWorld Award)
Automated
ToxCheck and
other CIx tools
SWISS-MODEL and 3D-Crunch recognised as a leading new concept (ComputerWorld Award)
UD recognised for visionary use of information technology in the category of Medicine (ComputerWorld Award)
EGEE‘06 / M. Peitsch / Sept, 2006
Systems Biology
Study and Understand Biological Networks / “GRIDs ;-)”
...
“Omics”
Experiments
control
cyto
nuc
6
3
4
2
2
1
0
-2
0
0
1
2
3
4
5
-1
0
1
2
drug
time
3
4
2
2
1
0
-2
3
4
5
3
4
5
time
6
0
0
1
2
Mathematical
3
4
5
-1
0
1
time
2
time
Models
EGEE‘06 / M. Peitsch / Sept, 2006
5
0
-5
- 10
- 15
- 20
- 25
- 30
0
50
100
150
200
Influencing Biomolecular Processes
Target
Ligand
Drug
ACTIVE
INACTIVE
Target = enzyme, receptor, nucleic acid, …
Ligand = substrate, hormone, other messenger, ...
EGEE‘06 / M. Peitsch / Sept, 2006
Target finding:
Protein Kinase CK2 has roles in cell growth,
proliferation and survival.
Protein Kinase CK2 has a possible role cancer
and its over expression has been
associated with lymphoma.
Target validation:
To elucidate the different functions and roles of CK2 and confirm it
as a drug target for oncology, one needs a potent and selective
inhibitor.
Approach:
The problem was addressed by in silico screening (docking).
EGEE‘06 / M. Peitsch / Sept, 2006
Steve Dschmeissner / Science Photo Library
Our 1st PC Grid Success Story: Protein Kinase CK2
Inhibition
Virtual Screening by in silico Docking
> 400,000
Compounds
Docking
Process
and
Selection
of
possible
hits
< 10
Compounds
EGEE‘06 / M. Peitsch / Sept, 2006
Important results
Conclusion
We have identified a 7-substitued Indoloquinazoline
compound as a novel inhibitor of protein kinase CK2 by
virtual screening of 400 000 compounds, of which a dozen
were selected for actual testing in a biochemical assay. The
compound inhibits the enzymatic activity of CK2 with an IC50
value of 80 nM, making it the mostpotent inhibitor of this
“The associated
reported work
enzyme ever reported. Its high potency,
with clearly shows that large database
high selectivity, provides a valuable tool
for the study
of the
docking
in conjunction
with appropriate scoring and
biological function of CK2.
filtering processes can be useful in medicinal chemistry.
This approach has reached a maturation stage where it
can start contributing to the lead finding process. At the
time of this study, nearly one month was necessary to
complete such a docking experiment in our laboratory
settings. The Grid computing architecture recently
developed by United Devices allows us to now perform the
same task in less than five working days using the power
of hundreds of desktop PC’s. High-throughput docking has
therefore acquired the status of a routine screening
technique.”
EGEE‘06 / M. Peitsch / Sept, 2006
Peru
In silico DD for Dengue ( Talk by M. Posdvinec)
EGEE‘06 / M. Peitsch / Sept, 2006
Proteome Informatics (Talk by P. Hernandez)
Trypsin
Slide from M. Podvinec
Extract
Isolate
Relative Intensität
Digest
[KR]|{P}
m/z
LC column
+
+
+
Separate
EGEE‘06 / M. Peitsch / Sept, 2006
AcCN
AcOH
HPLC
Knowledge GRIDs
Data and Information complexity
Literature
Molecular Structure
Anatomy
& Clinical
Pathways
Raw data from instruments
40
30
20
10
1
S S L L E K G L D G A K K A V G G L G K L G K D A V E D L E S V G K G A V H D V K D V L D S V L
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
b28-D (y26)
100
2200
39
40
41
42
43
44
45
46
47
48
y39 -D9
3832.1
b42 - D
4174.9
3600
b45 - D
4503.2
4059.6
b38
3717.1
3321.5
2900
y35
3491.6
2981.5
y33
3094.3
3167.7
2623.8
y27
2795.8
2081.1
1838.1
1911.5
1652.2
y18
b30
b27
b23
Mass (m/z)
EGEE‘06 / M. Peitsch / Sept, 2006
38
y20 -D
y11
1500
37
b39-D
2495.6
2324.7
50
2209.3
% Intensity
b24-D (y22)
36
2738.9
y24 -D
35
3876.3
Genomics and Proteomics
34
4290.3
1
4300
[M+H]+
5000
Connecting the Knowledge Bodies (requirements)
 Intelligent integration of heterogeneous data to
enable “Seamless Navigation”:
 One-stop shop.
 Re-useable, in any Web and Office application.
 Intelligent, i.e. knows about biology, medicine,
chemistry, diseases, business, people, etc…
 On demand and easy to use.
 Configurable.
EGEE‘06 / M. Peitsch / Sept, 2006
Connecting the Knowledge Bodies (Components)
 Indexing of large heterogeneous data collections
(databases, full texts) to enable semantic expansion.
 Information Retrieval and Extraction, entity
recognition, semantic enrichment.
 Knowledge Map (navigating the conceptual network).
 Terminology Hub (thesauri and ontologies).
 Ontology-associated business rules.
EGEE‘06 / M. Peitsch / Sept, 2006
What entities constitute our Terminology?
 Chemical entities – IUPAC names, trivial names,
trade names, INNs, compound codes, ligands.
 Biological entities – targets, genes/protein,
modes of actions…
 Diseases, Indications, Side Effects,
Contraindications
 Institutions, Affiliations, People
 Geographic locations
…
EGEE‘06 / M. Peitsch / Sept, 2006
The Ultralink: Contextual Hyperlinking
 The Ultralink is an “intelligent” context-sensitive Hyperlink
created at run time.
 The Ultralink is a menu of links instead of a single link.
 This menu will only offers sensible actions/options based on a set
of rules attached to an ontology.
 The UltraLink allows the dynamic inter-connection of any piece of
text or information with any database, search engine and
application in the Knowledge Space.
 The UltraLink enables seamless information Navigation
EGEE‘06 / M. Peitsch / Sept, 2006
The Ultralink can be called from many applications:
e.g. Internet Explorer
Internet Explorer Integration
GPS Add-in
4
User requests
for analysis
Injection of
specific HTML
tags
2
Sends the
document for
analysis
Web Service (WSDL)
Web Page
1
3
Gets back
tagged parts
GPS Lexical Analysis Server Tools
EGEE‘06 / M. Peitsch / Sept, 2006
Terminology
Zoning
DocStructures
Lexical Extraction
Tagging
Meta-Rules
Tagged Document
MouseOver
Click
Color coding according to concept type.
In this example:
Yellow = Gene Name; Red = Institution
EGEE‘06 / M. Peitsch / Sept, 2006
BLAST Interface
EGEE‘06 / M. Peitsch / Sept, 2006
EGEE‘06 / M. Peitsch / Sept, 2006
Acknowledgements
Novartis
Thérèse Vachon
Martin Romacker
University of Basel:
CSCS
Olivier Kreim
Torsten Schwede
Marie-Christine Sawley
Uwe Plikat
Michael Podvinec
Peter Kunszt
Pierre Parisot
Jürgen Kopp
Sergio Maffioletti
Nicolas Grandjean
Rainer Pöhlmann
Arthur Thomas
Brigitte Charpiot
Konstantin Arnold
Jean-Marc von Allmen
Dominique Zosso
Daniel Cronenberger
Eric Vangrevelinghe
Vital-IT:
Pascal Afflard, Armin Widmer
Victor Jongeneel
Christian Bartels & Said Karfane
Bruno Nyffeler
Jan van Oostrum & Team
Heinz Stockinger
Carolyn Cho & Team
EGEE‘06 / M. Peitsch / Sept, 2006