The role of chemoinformatics in the design of a comprehensive drug ...

The Role of Chemoinformatics in the Design of a
Comprehensive Drug Discovery Screening Collection
Edgar Jacoby
Center for Proteomic Chemistry
2nd Strasbourg Summer School on Chemoinformatics
VVF Obernai, France, 20-24 June 2010

Overview
‣ NIBR compound collection enhancement project
‣ External compound selection process
‣ Combinatorial library synthesis – scaffold projects
‣Compilation of annotated knowledge-based sets
‣Molecular Information System for pharmaceutical ligands
‣Knowledge-based ligand design and virtual screening strategies
‣Chemoinformatics: Quo Vadis
| Obernai| 20.06.10 | Edgar Jacoby

Essential Properties of Compounds
| Obernai| 20.06.10 | Edgar Jacoby

Essential Properties of Compounds - Continued
| Obernai| 20.06.10 | Edgar Jacoby
Jacoby et al. Chemogenomics 2007, 1-38

Externally Available Chemistry Space
LDCStructuresDB
‣Oracle and ISIS databases in
place
‣10.000.000+ Unique structures
‣13.000.000+ Catalogue entries
‣130+ Collections
‣50+ Vendors
LDCScaffoldDB
‣ Oracle and ISIS databases
‣ 18.000+ Unique scaffolds
‣ 15 Vendors
| Obernai| 20.06.10 | Edgar Jacoby

Chemistry, Biology, Informatics Based Compound Selection
0. The existing
archive
descriptor space
Similar to Know MDDR
Actives and Drugs
1. Select from
container 1:
(target family
related affinity)
high density,
similarity

Substructure/Fragment Filters
Definition of the Exclusion-Criteria:
A: Don't store in MCS
B: Don't produce as solution or store in SOLAR
C: Don' buy (but take for free)
ID Structure Chemical Description
33 A
Archive Exclusion Flag
Toxic
Unintersting Uninteresting
Reactive
O carbonic acid nitriles Y Y N
N
34 A
O O
* O *
*
O O
S
O
*
*
O O
P
O
*
acyclic acid anhydrides including
phoshonates and phosphinates, but
not di- and triphospates
Y Y N
O O O O
O O
P
O
P
S
* S
O P
O S
* * *
*
not:
O
O
P
O
O
P
O
O
O
35 B
O O
* O *
O O
S
* O *
O O
P
* O *
cyclic acid anhydrides including
phoshonates and phosphinates, but
not di- and triphospates
Y Y N
O O
P
O
P
*
O O
S
O S
* *
O O
S P
* O *
not:
O
O P O
O
O
P O
O
36 C H O
acyclic alkyl carbamates Y N N
H
O
N
| Obernai| 20.06.10 | Edgar Jacoby

In Silico Substructure Tox Alerts
| Obernai| 20.06.10 | Edgar Jacoby

Cheminformatics Platform for Selection and Logistics
Sample/catalogue entry domain
Structure domain
N
Load Catalogue, add new structures (new), check
availability for all compounds
Y
Export new structures for evaluation (exported)
N
Y
Catalogue entry marked
Catalogue entry
as rejected
marked as selected
Y
N
Structure based
evaluation of
compounds
Drug-Likeness and
Target Focus
Diversity
Compound is not
available
N
Don’t order (cancelled)
N
Confirm availability and
price
Y
Decide to order
Y
Order placed (ordered)
Structures processed
in parallel acquisition
processes
not registered
in corporate structure DB
Diversity
N
Identity by structure (smiles)
Compound received
Y
Register compounds
(registered)
registered
in corporate structure DB
Structures of “archive
compounds”
Schuffenhauer, A. et al. Comb. Chem. High Throughput Screen. (2004), 7, 771-782.
| Obernai| 20.06.10 | Edgar Jacoby

Scaffold Tree Example for HTS Results
PubChem Pyruvate Kinase Data Set
Schuffenhauer et al. J Chem Inf Model. 2007, 47, 47-58.
Color intensity by
fraction of actives
| Obernai| 20.06.10 | Edgar Jacoby

Evaluation Scheme for Combinatorial Hit Discovery Libraries
2004
Patentability
Novelty of Scaffold
Novelty of Products
Modularity
Preprototype Criteria
Planned Size
Ratio of Lowest and
Highest Nr. of BBs
Complementarity
Library Evaluation
Rule of Five
InSilico Tox
BB Availability
BB Diversity
Chiral Purity
HT Solubility
HT logP
HT Permeability
Prototype Criteria
1 1 1 0 0 -1 1 1 1 1 1 1 0 0 0 0 0 -1 0 1 1
CYP 450 (Metabolism
HT HERG channal
assay
Ames test
Ease of Realisation
Time of Development
Chemical Stability
1
0
-1
2007
novelty
target
focus
synthetic
complexity
potential
for HLO
computed
properties
Jacoby et al. Curr Top Med Chem. 2005, 5, 397-411
| Obernai| 20.06.10 | Edgar Jacoby

The Ideal Size of Combinatorial Libraries
| Obernai| 20.06.10 | Edgar Jacoby

Monitoring of Compound Collections - Lipinski Profiles
12.0%
AlogP Distribution
40.0%
H-Bond Donors
35.0%
10.0%
30.0%
% of samples
8.0%
6.0%
4.0%
CHC [%]
SOLAR [%]
CMedC [%]
% of samples
25.0%
20.0%
15.0%
CHC [%]
SOLAR [%]
CMedC [%]
10.0%
2.0%
5.0%
0.0%
-5 -4 -3 -2 -1 0 1 2 3 4 5 6 7 8 9 10
AlogP
0.0%
0 1 2 3 4 5 6 7 8 9 10
Count
10.0%
Molecular Weight Distribution
25.0%
H-Bond Acceptors
9.0%
8.0%
20.0%
% of samples
7.0%
6.0%
5.0%
4.0%
CHC [%]
SOLAR [%]
CMedC [%]
% of samples
15.0%
10.0%
CHC [%]
SOLAR [%]
CMedC [%]
3.0%
2.0%
1.0%
5.0%
0.0%
0 100 200 300 400 500 600 700 800 900 1000
Molecular Weight
| Obernai| 20.06.10 | Edgar Jacoby
0.0%
0 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15
Count

Prediction of Water Solubility
10.0
-logS n (exp, equilibrium)
9.0
8.0
7.0
6.0
5.0
4.0
3.0
2.0
logS vs AlogP, no salt
logS vs AlogP, salt
1.0
GSE Limit -logS >= logP + 0.5
Limit -logS >= logP
0.0
0 2 4 6 8 10
AlogP
-logS pH6.8(exp,equilibrium)
10.0
9.0
8.0
7.0
6.0
5.0
4.0
3.0
2.0
1.0
inside model range
outside model range
salts
exact
-1 log unit
+1 log unit
0.0
0.0 1.0 2.0 3.0 4.0 5.0 6.0 7.0 8.0 9.0 10.0
-logS pH6.8(QMPRPlus, Corina 3D structure)
10
10
-logS n (exp, equilibrium)
9
8
7
6
5
4
3
2
1
-logS pH6.8(exp, equilibrium)
9
8
7
6
5
4
3
2
1
151
144
147
84
148
85
149
118
146
150
153
154
152
143
145
142
142
139
112
90
140
141
137
138
76
108
109
135
134
133
136
126
127
130
129
131
128
2
132
4
1
non_salt
salt
exact
+1 log unit
-1 log unit
0
0 1 2 3 4 5 6 7 8 9 10
-logS n (QikProp, Corina 3D structures)
0
0 1 2 3 4 5 6 7 8 9 10
-logS pH6.8(calculated,ACD)
| Obernai| 20.06.10 | Edgar Jacoby

The “Egan Egg”: Statistical Prospectively Useful Model
8
Poor Solubility
Access Egan Egg via IsisDraw
InSilico Profile T2ABS plot
ClogP
6
4
2
0
-2
Higher
Permeability
• Good Hunting Grounds
Higher
Solubility
Poor Permeability
& Poor Solubility
• Few marketed drugs lie
in this region
Poor Permeability
• Drugs in this
region usually
work via prodrugs
of active transport
-20 0 20 40 60 80 100 120 140 160 180 200
| Obernai| 20.06.10 | Edgar Jacoby
PSA WJ Egan et al, JMC, 2000

Chemist Triage Score and Naïve Bayesian Models
‣ Yes/No decision from five individual triaging exercises taken to build one
Bayesian models
‣ All five models are used in a consensus score, fraction of models voting
against the molecules
Chemist Triage Score
100%
90%
16%
Histogram of NB Score
Screening Collection
Less attractive cpds.
Fraction of compounds
80%
70%
60%
50%
40%
30%
20%
10%
0%
MCS SLIA ICLF PSEB MCS
external
2003
MCS
external
2004
Drugs NAPR HTS Hits
2004
Random
Chemist
Triage
Score
Serie
1
s6
Serie
0.8
s5
Serie
0.6
s4
Serie
0.4
s3
Serie
0.2
s2
Serie
0
s1
% of samples
14% Primary Hit-list
12%
Training Set
10%
8%
6%
4%
2%
0%
-100 -80 -60 -40 -20 0 20 40 60 80 100
| Obernai| 20.06.10 | Edgar Jacoby

Examples of Unwanted Molecules
ChemistTriageScore: Excluded
N
H
O
1.0
N
H
O
NH 2
O
N H
N
O
H
N
Br
N +
Similog density: Excluded
HO
HO
N
H
0.0012
N
H
OH
OH
N + O OH
Cl
Cl
H
N
Cl
Cl
O
N +
O
O
N
F
F
F
O
HN
O
OH
0.8 0.8
0.0017
ChemistTriageScore: Penalized
Similog density: Penalized
O
O
N
HO
S
HO
HO
0.4
Cl
N
NH
Cl
N
O
HO
S
O
OH
HO
0.6
F
F
F
0.6
F
0.0029
O
O N
0.0036
N
| Obernai| 20.06.10 | Edgar Jacoby

The Challenge of Quality in Candidate Optimization
Biller S. et al. Biotechnology: Pharmaceutical Aspects, 2004, 413-429.
| Obernai| 20.06.10 | Edgar Jacoby

Safety Pharmacology Prediction
| Obernai| 20.06.10 | Edgar Jacoby

Which Annotated Compound Libraries Make Sense
1. Approved drugs – FDA orange book - SOSA
2. Other reference compounds (USAN, INN) with known bioactivity (Tocris,
Prestwick, etc.)
3. Compounds processed by profiling
4. Primary metabolites, natural products and derivatives (KEGG, etc.)
5. Target family focused libraries
6. Diversity selection – DOS / BIOS
7. Protein mimetics library: β-turn/α-helix mimetics
8. Fragment library for NMR/Xray screening
| Obernai| 20.06.10 | Edgar Jacoby

Focusing : Optimizing the Overlap Between Chemical and
Biological Spaces
Wess DDT 2002, 7, 533-535
| Obernai| 20.06.10 | Edgar Jacoby

Knowledge-Based Virtual Screening
Ligand-based
T
A
R
G
E
T
Yes
Annotation No
Yes
Known
Actives
No
No
Known
Structure
Yes
Yes
Are
Actives
‘Drug like’
No
No
Is binding
site ‘Drug
like’
Yes
Yes
Select
compounds
similar to
known actives
Select
compounds
compatible to
binding site
Similarity
based list
Diversity
screening
Structure
based list
Structure-based
| Obernai| 20.06.10 | Edgar Jacoby

Molecular Information System for Pharmaceutical Ligands of
the Main Target Families
Target
MIS
Reference
Ligands
Compound
Selections
Ligand -Target Sequence
Ligand -Target Classification
Ligand - Interaction Classification
Ligand - Functionality
Ligands
Novartis
MDDR,...
Targets
Swiss-Prot
GPCRDB,...
Schuffenhauer, A. and Jacoby, E. BioSilico ( 2004), 2, 190-200
| Obernai| 20.06.10 | Edgar Jacoby

Databases of Known Pharmaceutical Ligands and Drugs
Structure
MDDR-3D 2001.1
(23.01)
Extreg 185987
Pref#
Preview
Phase
185987
Act.invest
Launched
Index Activity
28000 Cardiotonic
31000 Antihypertensive
31432 Angiotensin II
AT1 Antagonist
Model
Structure
O
Identification
O
N
O
Biology
N
N
Lit & Patent
Chiral
N
N
Source
Novartis
Company.code
CGP-48933
CGP-63170 (comb. with
Generic name
Valsartan < Rec INN; USAN >
Trademark
Valpression (Menarini, IT)
Provas (Schwarz, DE)
Varexan (Egis, HU)
Di (N ti CA DE
| Obernai| 20.06.10 | Edgar Jacoby

Knowledge Based Approaches : GPCR I
MDL Patent Database
Activity Based Search
38 Patent Families
1810 Reference Structures
Similarity Search
Naïve Bayes Model
542 SOLAR in silico hits
| Obernai| 20.06.10 | Edgar Jacoby

Similarity Searching using Similog Keys
P. Floersheim’s Similog keys
counts of DABE triplets in 2D space
6
0100
O
0010
O
4
O
1100
H
6
0010-4-1100-6-0100-6-
1NN 3NN avg cent
Schuffenhauer, A. et al. J. Chem. Inf. Comp. Sci. ( 2003), 2, 391-405.
| Obernai| 20.06.10 | Edgar Jacoby

Design of TAM Libraries for Monoamine-Related GPCRs
Known Reference Architectures
Novel Prototypes
N H 2
HO
NH
Serotonin
5HT ago.
HO
H
N
O
Propranolol
β antago.
HO
N
8-OH-DPAT
5HT1A part. ago.
N
Cl
NH
O
N
N
H
N
N
NH
O S O
Cl
O
O
N
N
NH
O
F
Ketanserin
5HT2A antago.
HN
Cl
N
N
S
N
Cl
Janssen_1
D4 antago.
N
N
O
N
O
N
WAY-100635
5HT1A antago.
HN
O
N
N
NH
O S O
Cl
N
N
N
N
Cl
NH
N
N
O
5HT7 antago.
pKB = 7.25
O
O
NH 2
O
N
Cl
HO
H
OH
N
OH
N
N
RO-16814
β ago.
N
H
Kissei_1
D2 antago.
NH
O S O
Cl
O
O
Jacoby, E. Quant. Struct. Activity. Relations. (2001) 20, 115-123.
| Obernai| 20.06.10 | Edgar Jacoby

Deconvolute and recombine : Drug like molecules
Currently applied knowledge-based strategy
New Leads + New Affinity Profiles
Recompose
Databases of Active Fragments
Deconvolute
NOVARTIS DB
CAS DB
Jacoby et al. Drug News Perspect. (2003),16, 93-102.
| Obernai| 20.06.10 | Edgar Jacoby

Ligand selectivity in a subfamily with conserved molecular
recognition
pKi
α2A α2B α2C β1 β2 D2.L D3 D4.2 5HT1A 5HT1D 5HT1E 5HT1F 5HT2A H1
Sumatriptan < 6 < 6 < 6 < 6 < 6 < 6 < 6 < 6 6.4 8.4 < 6 7.8 < 6
Alniditan 7.4 6.8 7.7 < 6 < 6 < 6 6.5 7.1 8.4 9.4 6.6 6.4 < 6 < 6
Haloperidol < 6 6 .3 6.2 < 6 < 6 8.7 8.1 7.9 < 6 6.6 < 6 < 6 6.7 6.1
Pipamperone 6.1 7.5 6.5 < 6 < 6 6.9 6.6 8.3 < 6 6.8 < 6 < 6 8.3 < 6
Clozapine 7.3 7.7 8.1 < 6 < 6 6.7 6.6 7.4 6.9 6.4 6.4 6.9 8.0 9.6
Olanzapine 6.3 6.7 6.7 < 6 < 6 7.5 7.3 7.6 < 6 6.2 < 6 6.5 8.6 9.2
Do we understand and master the principles of profile composition in
terms of molecular recognition
| Obernai| 20.06.10 | Edgar Jacoby

Focus of TAM library – Similarity Histograms
6
5
Monoamine GPCRs
R1 R1'
N
R2
4
3
2
1
0
0
0.1
0.2
0.3
0.4
0.5
0.6
0.7
0.8
0.9
1
% Compounds
Peptide-binding GPCRs
LGIC
Maximum Tanimoto Index
| Obernai| 20.06.10 | Edgar Jacoby

A general strategy for creating “inactive” conformation kinase
inhibitors - ABL
Okram et al. Chemistry & Biology 2006, 13, 779–786
| Obernai| 20.06.10 | Edgar Jacoby

Selectivity Analysis of Combinatorial Libraries
100%
80%
60%
40%
20%
0%
BCG-001
BET-001
BET-02A
DIA-002
IMO-001
AAL-001
BET-02B
HOP-002
HOP-003
DIA-001
PUR-003
SYK-001
AAL-002
BZI-001
GUA-003
IMI-251
OXI-001
SRC-001
TAM-003
IAB-451
IPO-825
PUR-001
TAM-001
TAM-002
TSA-003
TAM-004
TAM-005
TSA-001
Is Target Family Focus a Myth
GPCR
Kinase
Protease
other Enzymes
NUR
ion channel
PPI
functional
| Obernai| 20.06.10 | Edgar Jacoby

Extend Knowledge-based Approaches
• Target family based approaches
• BIOS – Protein domain based designs
• Privileged Scaffolds
• Protein Secondary Structure Mimetics: α-helix/β-turn
• Co-factor mimetics: ATP, NADP, FAD, Co-A, SAM, etc.
| Obernai| 20.06.10 | Edgar Jacoby

Privileged structures / Scaffold target matrix
O
O
O
N
N
N
N
N
N
O
Naftopidil
α 1
antagonist
HO
O
O
O
H
N
O
N NH H N
N
N
O
N
N
N
H
N
Cl
N
N NH
Diovan
Meiji Seko
AT 1
antagonist µ 1
agonist / D 2
antagonist
H
O N
O
H
O
O
H
N
N
NH 2
N
O
O O
N
N
H
N
N
N
N
O
Merck_1
NK 1
antagonist
N
H
OH
5HT1A
5HT1B
5HT1D
5HT1F
5HT2A
5HT2B
5HT2C
5HT4
A1A
PS1 0 0 0 0 2 0 0 0 0 4 0 0 1 11 0 0 0 0 0 4
PS2 0 0 0 0 0 1 1 0 0 0 0 0 0 2 0 14 0 0 0 4
PS3 6 0 0 0 3 3 3 0 0 0 0 0 0 0 0 0 0 0 0 4
PS4 2 0 0 0 1 1 1 0 0 0 0 0 0 0 0 0 0 0 0 4
PS5 2 0 0 0 0 0 0 0 1 0 0 0 0 2 1 0 0 0 0 4
PS6 0 0 0 0 0 0 0 0 1 1 1 0 0 0 0 3 0 0 0 4
PS7 2 0 0 0 0 0 0 0 0 0 0 0 1 1 0 1 0 0 0 4
N
H
NH
N
N
A2A
B1
B3
D1
D2
PS1 PS2 PS3
PS4
N
N
D3
N
N
D4
M1
M3
H2
Frequency
Servier_1
NK 1
antagonist
Merck_2
GHr agonist
RS504393
α 1A
/ CCR 2b
antagonist
Pfizer_1
CCK antagonist
N
NH
O
O
O
HN
O
N
H
N
PS5 PS6 PS7
| Obernai| 20.06.10 | Edgar Jacoby

Privileged structures GPCRs
The 2-aryl-indole case
| Obernai| 20.06.10 | Edgar Jacoby

Co-factors
Bioactive Motifs with Conserved Binding Sites in Enzymes
| Obernai| 20.06.10 | Edgar Jacoby

Ligands vs. domains and folds that bind them
Analysis of PDB – Cofactor Binding Sites Open Large Potential Target Space
Ligands Full name of ligands Number of
domains
Number of
folds
PLP Pyridoxal-5'-phosphate 17 10
TDP ThiaminDiphosphate 36 10
SAM S-adenosylmethionine 41 13
COA CoenzymeA 29 14
GTP Guanosine-5'-triphosphate 26 15
AMP AdenosineMonophosphate 29 15
NAP NadpNicotinamide-adenine-dinucleotidePhosphate 73 16
FMN FlavinMononucleotide 122 16
NDP NadphDihydro-nicotinamide-adenine-dinucleotidePhosphate 44 18
ANP PhosphoaminophosphonicAcid-adenylateEster 59 19
FAD Flavin-adenineDinucleotide 152 21
NAD Nicotinamide-adenine-dinucleotide 149 27
GDP Guanosine-5'-diphosphate 86 29
ADP Adenosine-5'-diphosphate 137 31
ATP Adenosine-5'-triphosphate 97 35
Distribution patterns of small-molecule ligands in the protein universe and implications for origin of life and drug discovery
Hong-Fang Ji , et al.
Genome Biology 2007, 8:R176doi:10.1186/gb-2007-8-8-r176
| Obernai| 20.06.10 | Edgar Jacoby

Sam: S-Adenosylmethionine
Jack of all Trades and Master of Everything
N H 2
OH
O
S +
HO
O
HO
N
N
N
N
NH 2
Conformational Analysis of 252 SAM PDBs
3 main Clusters – Pharmacophorfamilies
(J. Priestle MLI)
Loenen WA.
Biochem Soc Trans. 2006 Apr;34(Pt 2):330-3
| Obernai| 20.06.10 | Edgar Jacoby

Knowledge-Based Virtual Screening – MDM4/2
Compounds From In-house MDM2 Assays and Similars
‣ Hit-finding and Hit-to-lead assays - compounds with IC50 < 50 µM (650)
‣ FCFP4 similarity search (1873)
‣ Hopfen QSAR (5000)
Literature and Patent Mining
‣ WOMBAT Literature Data and 14 Patent families from MDL Patent Database
‣ FCFP4 similarity search (675) and NB model (5000)
‣ FEPOPS (1000) similarity search, Substructure search (1707)
HTD
‣ Homology model based on SPIRO-OXINDOLE and NUTLIN MDM2 structures
‣ Constrained SP Glide docking
‣ Selection of top 20000 Compounds
UNITY Searching
‣ Homology model based on SPIRO-OXINDOLE and NUTLIN MDM2 structures
‣ Constrained search: Hydrophobic centers and HBonds (49305)
‣ Constrained SP Glide docking
‣ Selection of top 20000 Compounds
M
D
M
4
H
I
T
S
E
T
4
6
5
1
1
| Obernai| 20.06.10 | Edgar Jacoby

FEPOPS 3D similarity searching
N H 2
N
N
N
ATP
O P
O O
N
HO
OH
O
OH
O
OH
O
P OH
P
O
OH
HO
Balanol PKC inhibitor
O
H
N
O
N
H
O
O
OH
OH
HO
O
O
4 feature centroids and distances
2 4 2 4
Balanol low 2D similarity to ATP
High similarity in FEPOPS as in Xray
Computational Procedure
1. Generate conformers
2. Calculate FPs and properties
3. Compute fingerprints
4. Search based on fingerprints
Jenkins et al. J. Med. Chem. (2004), 47, 6144-6159.
| Obernai| 20.06.10 | Edgar Jacoby

Pharmacophore Screening – MDM4/2
Cl
H
O
N
H
O
N
Cl
O
N
NH
O
O
N
O
H
NH
H
HN
O
Cl
H
F
NH
Cl
| Obernai| 20.06.10 | Edgar Jacoby

Knowledge-Based Virtual Screening – MDM4/2
| Obernai| 20.06.10 | Edgar Jacoby

Comparing HTS and Virtual Screening – MDM4/2
Virtual Screen
HTS
| Obernai| 20.06.10 | Edgar Jacoby

Comparing HTS and virtual screening – MDM4/2
| Obernai| 20.06.10 | Edgar Jacoby

Chemoinformatics: Quo vadis
Towards prediction paradise
Van de Waterbeemd, H. et al. Nat Rev Drug Discov. 2003,192-204
| Obernai| 20.06.10 | Edgar Jacoby

Chemoinformatics: Quo vadis
Molecular Informatics
Field of
Informatics
Primary
Data
Similarity
Relationships
Key
Applications
Bio-
Protein
Sequences
Sequence
Alignments
Function Inference
Structure Determination
Protein Structures
Structure
Alignments
Function Inference
Structural
Site Annotation
Binding Sites
Site
Alignments
Target Hopping,
Cross Reactivity,
Selectivity,
Opportunity Mining
Sites+Ligands
Binding Mode
Alignments
Enhanced Screening,
Scaffold Hopping,
Novel Scaffolds
Small Molecule Docking
Chem-
Small Molecules
Molecular
Similarities
Screening
Courtesy of Dr. D. Debe, Eidogen-Sertanty
| Obernai| 20.06.10 | Edgar Jacoby

Chemoinformatics: Quo vadis
Chemical Systems Biology – Compound-pathway maps
| Obernai| 20.06.10 | Edgar Jacoby

Acknowledgements
• Ansgar Schuffenhauer
• Maxim Popov
• Kamal Azzaoui
• Joerg Muehlbacher
• Jeremy Jenkins
• Meir Glick
• Bill Egan
• Peter Ertl
• Bernard Faller
• Hans-Joerg Roth
• Peter Fuerst
| Obernai| 20.06.10 | Edgar Jacoby

Recent Publications
| Obernai| 20.06.10 | Edgar Jacoby