77
AcurePharma Bioinformatics, Metabolomics and personalized medicine Torbjörn Lundstedt KI 080609
AcurePharma
Bioinformatics, Metabolomics and personalized medicine
Torbjörn Lundstedt
KI 080609
AcurePharma 2/77CONFIDENTIALFigures from
http://www.nobel.se
Multivariate analysis of G-protein coupledreceptors
AcurePharma 3/77CONFIDENTIAL
A B
C
D
Bioinformatic examples
• G-proteincoupled receptors 2D structure GPCR
• 7 trans-membrane regions 7TM
• Amino acid sequences
AcurePharma 4/77CONFIDENTIAL
20 natural amino acids
• 20 natural amino acids
• Characterised with bothexperimental and calculated variables
• Observations = 20
• Variables = 26 (in this example)
• Have been expanded to 87 amino acids(Maria Sandberg et al.)
N = 20
K = 26
N = 87
AcurePharma 5/77CONFIDENTIAL
-4
-2
0
2
4
6
-4 -3 -2 -1 0 1 2 3 4 5 6 7
t[2
]
t [1]
87mia_all-020522.M2 (PCA-X), 20 natural mias data
t[Comp. 1]/t[Comp. 2]
ALA
ARG
ASN
ASP
CYS
GLN
GLU
GLY
HIS
ILELEU
LYS
METPHE
PROSER
THR
TRP
TYR
VAL
-0.20
-0.10
0.00
0.10
0.20
0.30
-0.20 -0.10 0.00 0.10 0.20 0.30
p[2
]
p[1]
87mia_all-020522.M2 (PCA-X), 20 natural mias data
p[Comp. 1]/p[Comp. 2]
MW
TL1
TL2TL3
TL4
TL5
TL6TL7
vdWvol
NM1NM7NM12
logP
EHOMO
ELUMO
HOF
POLAR
EN
Hardness
Stot
Spol
Snonpol
HDONR
HACCR
Chpos
Chneg
20 natural amino acids
• Two first principal components t1 vs. t2 � comparewith p1 and p2 for comparision
AcurePharma 6/77CONFIDENTIAL
-4
-3
-2
-1
0
1
2
3
4
5
-3 -2 -1 0 1 2 3 4 5 6 7
t[2]
t[1]
87mia_all-020522.M2 (PCA-X), Untitled
t[1]/t[2]
ALA
ARG
ASN
ASP
CYS
GLN
GLU
GLY
HIS
ILELEU
LYS
MET
PHE
PROSER
THR
TRP
TYR
VAL
20 natural amino acids
NH2
NH
O
OH
NH2
OHO
HO
H2N
O
OH
NH2
SO
OH
NH2
NH
NH
NH2O
OH
NH2HN
N
O
OH
NH2
NH2O
OH
NH2 O
NH2
O
HONH2
O
H2N
O
OH
NH2 O
OH
O
HO
NH2
O
HO
O
OH
NH2
OH
O
HO
H2N
O
OH
NH2
OH
O
HOH2N
O
OHH2N
O
HO
HN O
OH
AcurePharma 7/77CONFIDENTIAL
Principal properties natural amino acids
• Amino acids characterisedwith experimental and calculated variables
• PCA generates threedescriptors for each aminoacidz1 high lipofilic, lowhydrofobicz2 high large, low smallz3 elektronic properties
No. Name Name One letter code z1 z2 z3
1 Alanine ALA A 0.07 -1.73 0.09
2 Valine VAL V -2.69 -2.53 -1.29
3 Leucine LEU L -4.19 -1.03 -0.98
4 Isoleucine ILE I -4.44 -1.68 -1.03
5 Proline PRO P -1.22 0.88 2.23
6 Phenyalanine PHE F -4.92 1.30 0.45
7 Tryptophan TRP W -4.75 3.65 0.85
8 Methionine MET M -2.49 -0.27 -0.41
9 Lysine LYS K 2.84 1.41 -3.14
10 Arginine ARG R 2.88 2.52 -3.44
11 Histidine HIS H 2.41 1.74 1.11
12 Glycine GLY G 2.23 -5.36 0.30
13 Serine SER S 1.96 -1.63 0.57
14 Threonine THR T 0.92 -2.09 -1.40
15 Cysteine CYS C 0.71 -0.97 4.13
16 Tyrosine TYR Y -1.39 2.32 0.01
17 Aspargine ASN N 3.22 1.45 0.84
18 Glutamine GLN Q 2.18 0.53 -1.14
19 Aspartic acid ASP D 3.64 1.13 2.36
20 Glutamic acid GLU E 3.08 0.39 -0.07
Hellberg et al., 1987 (20 aa´s)Jonsson et al. 1989 (55 aa´s)Sandberg et al. 1998 (87 aa´s)
AcurePharma 8/77CONFIDENTIAL
Principal properties natural amino acids
Used to characterise amino acidsequences!
A V L
Principal
Property
translation
(z-scale)
A V L
z1 z2 z3 z1 z2 z3 z1 z2 z3
0.07 -1.73 0.09 -2.69-2.53-1.29 -4.19-1.03-0.98
AcurePharma 9/77CONFIDENTIAL
-20
-10
0
10
-10 0 10
t[2]
t[1]
7TM GPRCs analysis, t1/t2
897 sequences
675 variables
(135×5)
X
AcurePharma 10/77CONFIDENTIAL
-20
-10
0
10
-20 -10 0 10
t[4]
t[3]
gpa-s.M15 (PC), PCA all färg, Work setScores: t[3]/t[4]
Ellipse: Hotelling T2 (0,05)
Simca-P 8.0 by Umetrics AB 2001-09-06 12:58
7TM GPRCs analysis, t3/t4
AcurePharma 11/77CONFIDENTIAL
-0,100
-0,050
0,000
0,050
0,100
-0,100 -0,050 0,000 0,050 0,100
p[2
]
p[1]
gpa-s.M15 (PC), PCA all färg, Work setLoadings: p[1]/p[2]
Simca-P 8.0 by Umetrics AB 2001-09-26 08:39
A1t1
A1t2
A1t3
A1t4
A1t5A2t1
A2t2
A2t3A2t4
A2t5
A3t1
A3t2
A3t3A3t4
A3t5
A4t1
A4t2
A4t3
A4t4
A4t5
A5t1
A5t2
A5t3
A5t4
A5t5A6t1
A6t2
A6t3
A6t4
A6t5
A7t1
A7t2
A7t3A7t4
A7t5
A8t1
A8t2 A8t3A8t4
A8t5
A9t1A9t2A9t3A9t4
A9t5
A10t1
A10t2A10t3
A10t4
A10t5A11t1
A11t2
A11t3
A11t4A11t5
A12t1
A12t2A12t3
A12t4
A12t5
A13t1
A13t2
A13t3A13t4
A13t5
A14t1
A14t2
A14t3
A14t4
A14t5
A15t1A15t2
A15t3A15t4
A15t5
A16t1
A16t2A16t3
A16t4
A16t5
A17t1
A17t2
A17t3
A17t4A17t5
A18t1
A18t2
A18t3
A18t4A18t5
A19t1
A19t2
A19t3A19t4
A19t5
A20t1
A20t2 A20t3
A20t4
A20t5
B1t1
B1t2B1t3
B1t4
B1t5
B2t1
B2t2
B2t3B2t4
B2t5
B3t1
B3t2B3t3
B3t4
B3t5
B4t1B4t2
B4t3
B4t4
B4t5
B5t1B5t2
B5t3
B5t4
B5t5
B6t1B6t2
B6t3
B6t4
B6t5
B7t1
B7t2
B7t3B7t4B7t5
B8t1B8t2B8t3
B8t4B8t5
B9t1
B9t2 B9t3B9t4B9t5
B10t1
B10t2
B10t3B10t4
B10t5
B11t1
B11t2B11t3
B11t4
B11t5
B12t1
B12t2
B12t3
B12t4
B12t5
B13t1
B13t2
B13t3
B13t4
B13t5
B14t1
B14t2
B14t3
B14t4
B14t5B15t1B15t2 B15t3
B15t4
B15t5B16t1
B16t2
B16t3
B16t4
B16t5
B17t1 B17t2
B17t3 B17t4
B17t5
B18t1
B18t2
B18t3
B18t4
B18t5B19t1
B19t2
B19t3
B19t4
B19t5
B20t1
B20t2
B20t3
B20t4
B20t5
C1t1
C1t2
C1t3
C1t4C1t5
C2t1
C2t2
C2t3
C2t4
C2t5
C3t1
C3t2C3t3C3t4
C3t5
C4t1
C4t2
C4t3C4t4
C4t5
C5t1
C5t2
C5t3
C5t4
C5t5
C6t1
C6t2
C6t3
C6t4
C6t5C7t1
C7t2
C7t3C7t4
C7t5
C8t1
C8t2
C8t3
C8t4
C8t5
C9t1
C9t2C9t3
C9t4
C9t5
C10t1
C10t2
C10t3
C10t4C10t5
C11t1
C11t2
C11t3
C11t4
C11t5
C12t1
C12t2
C12t3
C12t4
C12t5
C13t1
C13t2
C13t3
C13t4
C13t5
C14t1
C14t2
C14t3
C14t4
C14t5C15t1C15t2C15t3C15t4
C15t5
C16t1 C16t2C16t3
C16t4C16t5
C17t1
C17t2
C17t3
C17t4
C17t5
C18t1C18t2
C18t3C18t4
C18t5
C19t1C19t2
C19t3
C19t4C19t5
C20t1
C20t2
C20t3
C20t4
C20t5D1t1D1t2
D1t3D1t4
D1t5
D2t1
D2t2
D2t3
D2t4
D2t5D3t1D3t2
D3t3
D3t4
D3t5D4t1
D4t2
D4t3D4t4D4t5
D5t1
D5t2
D5t3
D5t4
D5t5
D6t1
D6t2 D6t3
D6t4
D6t5D7t1
D7t2
D7t3
D7t4
D7t5 D8t1
D8t2
D8t3D8t4
D8t5D9t1
D9t2
D9t3
D9t4
D9t5
D10t1D10t2
D10t3
D10t4D10t5
D11t1D11t2
D11t3
D11t4D11t5
D12t1
D12t2
D12t3
D12t4
D12t5D13t1
D13t2
D13t3
D13t4
D13t5
D14t1D14t2
D14t3D14t4D14t5
D15t1
D15t2
D15t3
D15t4
D15t5D16t1
D16t2D16t3D16t4D16t5
D17t1
D17t2
D17t3
D17t4
D17t5
D18t1
D18t2
D18t3
D18t4
D18t5
E1t1
E1t2
E1t3
E1t4
E1t5 E2t1
E2t2 E2t3
E2t4
E2t5
E3t1
E3t2
E3t3
E3t4
E3t5
E4t1E4t2
E4t3E4t4E4t5
E5t1
E5t2
E5t3
E5t4
E5t5
E6t1
E6t2
E6t3
E6t4
E6t5
E7t1
E7t2
E7t3
E7t4
E7t5
E8t1
E8t2E8t3
E8t4
E8t5
E9t1E9t2
E9t3E9t4
E9t5E10t1
E10t2E10t3
E10t4E10t5
E11t1
E11t2
E11t3
E11t4
E11t5E12t1
E12t2E12t3
E12t4E12t5E13t1
E13t2E13t3
E13t4 E13t5
E14t1 E14t2
E14t3E14t4
E14t5
E15t1
E15t2
E15t3
E15t4
E15t5E16t1
E16t2E16t3E16t4
E16t5
E17t1
E17t2
E17t3E17t4
E17t5
E18t1
E18t2E18t3E18t4E18t5
E19t1
E19t2
E19t3 E19t4E19t5
E20t1E20t2
E20t3
E20t4E20t5
F1t1
F1t2
F1t3
F1t4
F1t5F2t1
F2t2F2t3
F2t4
F2t5
F3t1
F3t2
F3t3
F3t4F3t5F4t1
F4t2F4t3
F4t4
F4t5
F5t1
F5t2F5t3
F5t4
F5t5
F6t1
F6t2
F6t3
F6t4
F6t5
F7t1
F7t2
F7t3
F7t4F7t5
F8t1
F8t2
F8t3F8t4
F8t5F9t1
F9t2
F9t3
F9t4
F9t5
F10t1
F10t2
F10t3
F10t4F10t5
F11t1
F11t2
F11t3F11t4
F11t5
F12t1
F12t2
F12t3
F12t4
F12t5
F13t1
F13t2
F13t3F13t4
F13t5F14t1
F14t2
F14t3F14t4F14t5
F15t1
F15t2
F15t3
F15t4
F15t5
F16t1
F16t2
F16t3
F16t4
F16t5
F17t1
F17t2
F17t3
F17t4F17t5
F18t1
F18t2F18t3
F18t4F18t5
F19t1
F19t2
F19t3
F19t4
F19t5
G1t1
G1t2G1t3
G1t4G1t5
G2t1
G2t2
G2t3
G2t4
G2t5
G3t1
G3t2
G3t3G3t4
G3t5G4t1
G4t2G4t3G4t4
G4t5
G5t1
G5t2G5t3G5t4
G5t5
G6t1
G6t2
G6t3
G6t4
G6t5
G7t1
G7t2
G7t3
G7t4
G7t5
G8t1
G8t2
G8t3
G8t4
G8t5G9t1
G9t2
G9t3
G9t4
G9t5
G10t1
G10t2
G10t3
G10t4
G10t5
G11t1
G11t2
G11t3
G11t4
G11t5
G12t1
G12t2
G12t3
G12t4G12t5G13t1G13t2G13t3G13t4G13t5
G14t1
G14t2
G14t3G14t4 G14t5G15t1
G15t2 G15t3G15t4G15t5
G16t1G16t2G16t3G16t4
G16t5
G17t1
G17t2
G17t3
G17t4G17t5
G18t1
G18t2
G18t3G18t4
G18t5
Interpretation – difficulties
AcurePharma 12/77CONFIDENTIAL
Hierarchical PCA
• Dividing the data in different levels dependingon origin
• More than one PCA model to examine oneproblem
• Simplify interpretation
AcurePharma 13/77CONFIDENTIAL
Trans-membrane characterisation
Amino acid characterisation
NH2
NH
NH
NH2O
OH
H2N
O
OH
NH2
SHO
HOH2N
O
OH
NH2
NH
O
OH
NH2
O
OH
NH2
O
OH
NH2
NH2O
OH
HN O
OH
H2N
O
HO
NH2
SO
OH
NH2HN
N
O
OH
H2N
O
OH
NH2
OH
O
HO
NH2 O
NH2
O
HONH2 O
OH
O
HO
NH2
O
H2N
O
OH
NH2
O
HO
O
OHNH2
OH
O
HO
NH2
OHO
HO
X
Receptor characterisation
A(Taa) B(Taa) D(Taa) G (Taa)C(Taa) E(Taa) F (Taa)
A B C D E F G
Principal scores
Principal loadings
One receptor
level
Level 1 – InterpretationCharacterisation
PCA
AcurePharma 14/77CONFIDENTIAL
-4
-2
0
2
4
6
-5 -4 -3 -2 -1 0 1 2 3 4 5
t[2]
t[1]
amamamamamamamamamamamamamamamamamamamamamamamamamam
amamamamamamamamamamamamam
amamamamamamamamamamam amamamamamam
amamamamamamamamamamamamamamamamamamamamamamam
amamamamamamamamamamamamamamamamamamamamamamamamamamamamamamamam amamamamamam
amamam
am
amamamamam
amamamamam
amamamamamamamamamamamamamam
amamam
amamamamam amamam
amamamamamamam
amamamamamamamamamamam
amam
amamamamamamamamamamamamamamamam
amamamam
amam
amamamamam
amam
am
pepepepepepepepepepepepepepepe
pepepepepepepepepe pepepepepepe
pepepe pe pepepe
pe
pepepepepepepepepe
pe
pepepepepepepepepe
pe
pepepepepepepepepepe
pepepepepepepepepepepepepepepepepepepepepepepepepepepepepepe
pepe
pepepepepepepepepepepepe
pepepepepepepepepepepepepe
pepepepepepepepepepepepepepepepepepe
pepe
pepepepepepepepepepepepepepe
pepepe
pepe
pepepepepepepepepepe
pepepe
pepepepe
pepepepe
pepepe
pe pepepepe pepepe
pepepepe
pepepe
pe pe
pepepepepepepepepepepepepepepepepepepepepepe
pepepepepepepepepepepe
pepepe
pepepepepe
pepepepepepepepepe pe
pepe pepepepepe
pepepepepepepepepe
pepepepe
pepepepepepepepepepepe
pepepepepepepepepepe
pepepepepe
hphphphphphphphphphphphphphphphphphp
hphphphphp
opopopopopopop
opopopopopopopopopop
opopop
opopopop
opopopopopopop opopop
opopopopopopopop
opopopop
opop
opopopopopopopopopopopopopopop
opopopopopopopopopop
opopopopopop
opopopop
opopop
opopopopop
op op
opop
opopopopopop
op opopopop
op
opopopopopopopopop
op
op
opop
opop
opopopopop
opop opop
olololol
ol ololol
olol
ol ololololololololololololololol
olol
ololol
ololololol ol ol
ololol olololol
olol
ol ol olololol
olololololol
ololol
nunununununununununununununununu nununu
nununu nununu
nunununununununununununununu
nu
nunununu
nucbcbcbcbcbcb
cbcbcb papapapa
grgrgrgrgrgr
grtrtrtrtrtrtr trtrtrtrtr
trtrmlmlmlmlmlmlmlmlml
ml mlmlml ororor
oror
ororor orororororor oror
ororororor
orororor or
orororororor
oror
ororor
oror ororor
ororor
or
or
or
oror
ororor
oror
oror
ororororor
orororororor
or
ororororor
or oror
or oror
oror
Hierarchical PCA
-0,40
-0,30
-0,20
-0,10
0,00
0,10
0,20
0,30
-0,40 -0,30 -0,20 -0,10 0,00 0,10 0,20 0,30 0,40p[2
]
p[1]
At[1]
At[2]
At[3]
At[4]
Bt[1]
Bt[2]
Bt[3]
Bt[4]
Ct[1]
Ct[2]
Ct[3]
Ct[4]
Dt[1]
Dt[2]
Dt[3]
Dt[4]
Et[1]
Et[2]
Et[3]
Et[4]
Ft[1]
Ft[2]
Ft[3]
Ft[4]
Gt[1]
Gt[2]
Gt[3]
Gt[4]
AcurePharma 15/77CONFIDENTIAL
-6
-4
-2
0
2
4
-8 -7 -6 -5 -4 -3 -2 -1 0 1 2 3 4 5 6 7 8
t[2]
t[1]
s_hierar.M5 (PC), fyra komp., Work setScores: t[1]/t[2]
Ellipse: Hotelling T2 (0,05)
Simca-P 8.0 by Umetrics AB 2001-09-10 09:24
Analysis of sub-groups
-0,40
-0,30
-0,20
-0,10
0,00
0,10
0,20
0,30
0,40
-0,20 -0,10 0,00 0,10 0,20 0,30p[2
]
p[1]
s_hierar.M5 (PC), fyra komp., Work setLoadings: p[1]/p[2]
At[1]
At[2]
At[3]
At[4]
Bt[1]
Bt[2]
Bt[3]
Bt[4]
Ct[1]
Ct[2]
Ct[3]
Ct[4]
Dt[1]Dt[2]
Dt[3]
Dt[4]Et[1]
Et[2]
Et[3]
Et[4]
Ft[1]
Ft[2]
Ft[3]
Ft[4]
Gt[1]
Gt[2]
Gt[3]
Gt[4]
AcurePharma 16/77CONFIDENTIAL
Classification – GPCR
-10
0
10
-20 -10 0 10
t[2
]
t[1]
MC1O77616 MSHR_SHEEP
MSHR_CAPCA MSHR_RANTA
MSHR_CEREL MSHR_CAPHI
MSHR_ALCAA MSHR_BOVIN
MSHR_DAMDA MSHR_OVIMO
MSHR_HUMAN MSHR_MOUSE
MSHR_VULVU MSHR_CHICK1
MC2O57317
ACTR_HUMAN
ACTR_MESAU
ACTR_BOVIN
ACTR_MOUSE
MC3,4,5O93259 O73667
MC3R_HUMAN MC4R_HUMAN
MC4R_RAT MC3R_MOUSE
MC3R_RAT O73671
MC5R_BOVIN MC5R_MOUSE
MC5R_HUMAN MC5R_SHEEP
MC5R_RAT
Analysis of the melanocortin receptors; MC1, MC2, MC3, MC4 and MC5. MC1 and MC2 are quite different compared
to MC3-MC5.
AcurePharma 17/77CONFIDENTIAL
Classification – GPCR
5
0
-1
-10
-1
MC5MC5R_RAT
MC5R_MOUSE
MC5R_SHEEP
MC5R_BOVIN
MC5R_HUMAN
O73671
-5
0
5
0 10 20
t[2
]
t[1]
MC4MC4R_RAT
O73667
MC4R_HUMAN
MC3MC3R_RAT
MC3R_MOUSE
MC3R_HUMAN
O93259
- 0. 100
- 0. 050
0. 000
0. 050
0. 100
- 0. 100 - 0. 050 0. 000 0. 050 0. 100
p[2
]p[1]
7tm.M34 (PCA-X), PCA mc3,4,5
p[Comp. 1]/p[Comp. 2]
A1t1
A1t2A1t3A1t4A1t5
A3t1
A3t2
A3t3
A3t4A3t5
A5t1
A5t2A5t3A5t4A5t5
A6t1
A6t2
A6t3A6t4A6t5
A9t1
A9t2
A9t3
A9t4A9t5
A10t1
A10t2
A10t3
A10t4A10t5
A13t1A13t2A13t3A13t4
A13t5
A16t1
A16t2
A16t3
A16t4A16t5A20t1
A20t2
A20t3
A20t4
A20t5
B4t1
B4t2
B4t3
B4t4
B4t5B5t1
B5t2B5t3 B5t4
B5t5
B8t1
B8t2
B8t3B8t4B8t5
B9t1B9t2B9t3B9t4
B9t5
B16t1
B16t2
B16t3B16t4
B16t5
B18t1B18t2
B18t3
B18t4B18t5
B19t1
B19t2B19t3
B19t4
B19t5
B20t1
B20t2
B20t3
B20t4
B20t5
C3t1
C3t2C3t3C3t4
C3t5
C6t1C6t2C6t3
C6t4
C6t5
C8t1
C8t2C8t3C8t4C8t5
C9t1
C9t2C9t3C9t4C9t5
C12t1C12t2C12t3C12t4
C12t5
C14t1C14t2
C14t3
C14t4C14t5
C17t1
C17t2
C17t3
C17t4
C17t5
C20t1
C20t2
C20t3
C20t4C20t5
D1t1
D1t2
D1t3
D1t4
D1t5
D2t1
D2t2D2t3D2t4D2t5
D4t1
D4t2
D4t3
D4t4
D4t5
D5t1
D5t2D5t3 D5t4
D5t5
D8t1D8t2
D8t3 D8t4D8t5 D9t1
D9t2
D9t3D9t4
D9t5
D11t1D11t2
D11t3
D11t4
D11t5
D12t1
D12t2
D12t3
D12t4
D12t5
D13t1D13t2
D13t3D13t4
D13t5
D15t1
D15t2
D15t3
D15t4D15t5
D16t1
D16t2
D16t3D16t4
D16t5
E1t1
E1t2
E1t3
E1t4
E1t5
E5t1E5t2
E5t3E5t4E5t5
E8t1E8t2E8t3E8t4
E8t5
E9t1E9t2
E9t3E9t4E9t5E11t1
E11t2
E11t3E11t4
E11t5
E12t1
E12t2
E12t3E12t4
E12t5
E13t1E13t2E13t3E13t4
E13t5
E14t1E14t2E14t3E14t4
E14t5
E15t1E15t2
E15t3E15t4E15t5
E16t1E16t2E16t3E16t4E16t5
E19t1
E19t2
E19t3
E19t4
E19t5F1t1
F1t2F1t3 F1t4F1t5 F2t1F2t2
F2t3F2t4F2t5
F3t1F3t2F3t3F3t4
F3t5
F5t1
F5t2F5t3F5t4F5t5
F7t1
F7t2
F7t3
F7t4F7t5
F9t1
F9t2
F9t3F9t4
F9t5
F10t1
F10t2F10t3F10t4
F10t5
F13t1F13t2F13t3
F13t4F13t5
G3t1
G3t2
G3t3
G3t4G3t5
G9t1G9t2G9t3
G9t4G9t5
G10t1
G10t2
G10t3
G10t4G10t5
G11t1G11t2G11t3G11t4
G11t5
G18t1G18t2G18t3G18t4
G18t5
AcurePharma 18/77CONFIDENTIAL
Schematic Overview of MC1 and MC3 Receptor Chimeras
A B
C
D
Divided in four parts – characterised with binary coding
AcurePharma 19/77CONFIDENTIAL
Position 1 2 3 4 5 6 7 8 9 10 11 12 13
Seq. Part a b
Peptide
MSH Ser Tyr Ser Met Glu His Phe Arg Trp Gly Lys Pro Val
MS04 Ser Ser Ile Ile Ser His Phe Arg Trp Gly Leu Cys Asp
MS05 Ser Ser Ile Ile Ser His Phe Arg Trp Gly Lys Pro Val
MS06 Ser Tyr Ser Met Glu His Phe Arg Trp Gly Leu Cys Asp
Sequences of Melanocortin Receptor Active Peptides
AcurePharma 20/77CONFIDENTIAL
R e s p o n s e
Z X Y
R e c e p t o r P e p t i d e A B C D a b L o g ( K i )
A 1 B 1 C 1 D 1 a 1 b 1 0 0 0 0 0 0 0 . 2 1
A 1 B 1 C 1 D 1 a 2 b 2 0 0 0 0 1 1 0 . 8 2
A 1 B 1 C 1 D 1 a 2 b 1 0 0 0 0 1 0 - 0 . 0 7
A 1 B 1 C 1 D 1 a 1 b 2 0 0 0 0 0 1 1 . 6 5
A 1 B 2 C 2 D 2 a 1 b 1 0 1 1 1 0 0 1 . 3 7
A 1 B 2 C 2 D 2 a 2 b 2 0 1 1 1 1 1 3 . 8 2
A 1 B 2 C 2 D 2 a 2 b 1 0 1 1 1 1 0 2 . 0 3
A 1 B 2 C 2 D 2 a 1 b 2 0 1 1 1 0 1 2 . 7 2
A 1 B 1 C 1 D 2 a 1 b 1 0 0 0 1 0 0 1 . 3 7
A 1 B 1 C 1 D 2 a 2 b 2 0 0 0 1 1 1 3 . 3 6
A 1 B 1 C 1 D 2 a 2 b 1 0 0 0 1 1 0 2 . 3
A 1 B 1 C 1 D 2 a 1 b 2 0 0 0 1 0 1 2 . 1 8
A 1 B 2 C 2 D 1 a 1 b 1 0 1 1 0 0 0 1 . 5 6
A 1 B 2 C 2 D 1 a 2 b 2 0 1 1 0 1 1 4 . 1 4
A 1 B 2 C 2 D 1 a 2 b 1 0 1 1 0 1 0 2 . 3 1
A 1 B 2 C 2 D 1 a 1 b 2 0 1 1 0 0 1 2 . 8 3
A 1 B 2 C 1 D 1 a 1 b 1 0 1 0 0 0 0 1 . 8 5
A 1 B 2 C 1 D 1 a 2 b 2 0 1 0 0 1 1 4 . 1 9
A 1 B 2 C 1 D 1 a 2 b 1 0 1 0 0 1 0 2 . 5 7
A 1 B 2 C 1 D 1 a 1 b 2 0 1 0 0 0 1 3 . 7
A 1 B 1 C 2 D 2 a 1 b 1 0 0 1 1 0 0 1 . 5 5
A 1 B 1 C 2 D 2 a 2 b 2 0 0 1 1 1 1 3 . 1 1
A 1 B 1 C 2 D 2 a 2 b 1 0 0 1 1 1 0 2 . 4 1
A 1 B 1 C 2 D 2 a 1 b 2 0 0 1 1 0 1 2 . 4
A 1 B 1 C 2 D 1 a 1 b 1 0 0 1 0 0 0 0 . 5 4
A 1 B 1 C 2 D 1 a 2 b 2 0 0 1 0 1 1 1 . 8 4
A 1 B 1 C 2 D 1 a 2 b 1 0 0 1 0 1 0 0 . 8 6
A 1 B 1 C 2 D 1 a 1 b 2 0 0 1 0 0 1 1 . 5 7
A 2 B 2 C 1 D 1 a 1 b 1 1 1 0 0 0 0 1 . 7 7
A 2 B 2 C 1 D 1 a 2 b 2 1 1 0 0 1 1 3 . 0 8
A 2 B 2 C 1 D 1 a 2 b 1 1 1 0 0 1 0 2 . 7 6
A 2 B 2 C 1 D 1 a 1 b 2 1 1 0 0 0 1 2 . 8 8
A 2 B 2 C 2 D 1 a 1 b 1 1 1 1 0 0 0 1 . 8 4
A 2 B 2 C 2 D 1 a 2 b 2 1 1 1 0 1 1 3 . 6 7
A 2 B 2 C 2 D 1 a 2 b 1 1 1 1 0 1 0 2 . 6 3
A 2 B 2 C 2 D 1 a 1 b 2 1 1 1 0 0 1 2 . 8 6
A 2 B 2 C 2 D 2 a 1 b 1 1 1 1 1 0 0 1 . 9 7
A 2 B 2 C 2 D 2 a 2 b 2 1 1 1 1 1 1 4 . 4 9
A 2 B 2 C 2 D 2 a 2 b 1 1 1 1 1 1 0 3 . 0 2
A 2 B 2 C 2 D 2 a 1 b 2 1 1 1 1 0 1 3 . 1 7
Z
M o l e c u l e s D e s c r i p t o r s X Y
X
Receptor-Ligand-Response Matrix
AcurePharma 21/77CONFIDENTIAL
TM1
TM7 / 6
TM2 / 3 TM
4 /
5
A
B
C
D
ab
Interaction Between Peptide and MCR
AcurePharma 22/77CONFIDENTIAL
Biological samplesBiological samples
Biochemical analysis Biochemical analysis
of endogenous of endogenous
metabolitesmetabolites
Information?Information?The challenge in modern biology is not in data collection but rather in maximizing information in data -How to do it!
Metabonomics – Simplified view
AcurePharma 23/77CONFIDENTIAL
Metabonomics at a glanceNormal Disease Disease treated Treated healthy
Sample analysisData analysis
Classification
Metabolic profiles
AcurePharma 24/77CONFIDENTIAL
Strategy1. Formalize the aim
• What do we want?
2. Selection of objects• Design of Experiments (DOE)• Multivariate design (MVD)
3. Sample preparation and profiling of human and animal samples• In vivo, in vitro samples• Blood, Urine, Cerebral Spinal Fluid (CSF)• Synovial fluid (joint), Bowel fluids, Feces, Tissues
4. Integration and evaluation of collected data • Exploratory analysis, Interpretation & Visualization• Prediction models • Patterns• Target identification
5. Identify and define IPR opportunities and strategies
AcurePharma 25/77CONFIDENTIAL
Pigs
• Heart Ich/rep
• Microdialysis
AcurePharma 26/77CONFIDENTIAL
-1
0
1
-6 -5 -4 -3 -2 -1 0 1 2 3
t[2]
t[1]
SpacePig.M10 (PLS), Untitled, Work setScores: t[1]/t[2]
Ellipse: Hotelling T2 (0,05)
70 8090
100110120
130
140150
160
170180
190
200
210
220230
240250260
270
70
8090
100
110
120
130
140
150
160170 180
190
200210220
230
240250
260
270
Simca-P 7.01 by Umetri AB 2001-06-06 11:01
-0,40
-0,30
-0,20
-0,10
0,00
0,10
0,20
0,30
0,40
0,50
0,60
-0,40 -0,30 -0,20 -0,10 0,00 0,10 0,20 0,30
p[2
]
p[1]
SpacePig.M10 (PLS), Untitled, Work setLoadings: p[1]/p[2]
Adeno
Guano
Inos
Hypox
Lactat
Pyruvat
Nitrit
Nitrat
Urinsyra
MDA
Simca-P 7.01 by Umetri AB 2001-06-06 10:51
AcurePharma 27/77CONFIDENTIAL
”Data filtering”
• Large data – difficult to handle and overview
• Starting point – ask specific question and filter the data, use prior information
• Depending on the question the analysis can be focused at finding different patterns
• Information identified in data subsets can thereafter be applied to whole dataset
• Combine metabonomics data with questionnaire � connection genetics vs. lifestyle and environment
AcurePharma 28/77CONFIDENTIAL
Parameters ?• Example of filtering questions
• Compound classification – prior information
� Mechanism of action
� Classes of compound
� Indication
� Side effects
� Etc
• Use structure characterisation and PCA to generate principal properties � make representative/diverse selection to base first model on (e.g. sub-set)
AcurePharma 29/77CONFIDENTIAL
Multivariate design• Make PCA on reduced data set
• Make multivariate design(s) in the generated scores (principal properties) � cluster based design
• Select training set and test set according to a multivariate design – if possible
� Validate the model
� Obtain a more robust model
� Classify
� Identify biomarkers
AcurePharma 30/77CONFIDENTIAL
And combinations
there of
And combinations
there of
Example of questions for design
• Gender
� male/femaleLooking for genetic patterns and lifestyle patterns/markers separating the sexes
• Genetic profile� Asian, European, American
Looking for genetic patterns/markers
• Nationality
� China, UK or Asian, EuropeanLooking for lifestyle and environment patterns/markers
• Anamnesis
� ”healthy” vs. un-healthyLooking for disease pattern/markers
AcurePharma 31/77CONFIDENTIAL
Data layout
50 000Metabonomics
data Nationality
Male
Female Sick/healthy
Analysis in each filtering step (PCA)
With and withoutquestionnaire
AcurePharma 32/77CONFIDENTIAL
Data layout
Sick/healthy• Pregnant (folic acid, hormones)• Diabetes (insulin)• Gastric/stomach ulcer(Losec/Nexium)
• Head ace (paracetamole)• Smoker/non-smoker• Age < 50 <• BMI
AcurePharma 33/77CONFIDENTIAL
Lifestyle vs. metabonomics
• Identifying patterns in lifestyle by relating metabonomics to questionnaire (PLS)
50 000Metabonomics
dataQuestionnaire
AcurePharma 34/77CONFIDENTIAL
Parameters ?• Example of filtering questions
• Compound classification – prior information
� Mechanism of action
� Classes of compound
� Indication
� Side effects
� Etc
• Use structure characterisation and PCA to generate principal properties � make representative/diverse selection to base first model on (e.g. sub-set)
AcurePharma 35/77CONFIDENTIAL
Possible problems• Selection of control group in vivo?
• Selection of reference structures?
• Selection of colon's (LC/GC)?
• Sample handling/work up?
• Training set, test set – model validation, classification, biomarker
• Fast/Slow responders
AcurePharma 36/77CONFIDENTIAL
Class specific models 2 - PLS-DA• PLS-DA Maximum
separation projection
• PLS model built on group membership (1 or 0)
• Coefficients show as a whole how do the variables change when going from one group to another
• Here controls (f) and treated (fc) are separated
AcurePharma 37/77CONFIDENTIAL
PLS-DA between “f” and “fc”
• Coefficients show as a whole how do the variables change when going from one group to another
• Chemical shift regions or masses influential for the separation of the two classes
AcurePharma 38/77CONFIDENTIAL
Incision model short description
• Surgical intervention in rats. The SCI was induced by making a longitudinal incision into the right dorsal horn of the T10-11 segment
• AP173 and AP713 was given post-injury 5 minutes after injury with a topical dose of 5 µg in saline solution
• Tarlov scale was tested during 1 to 7 days
• The other responses was estimated day 7
AcurePharma 39/77CONFIDENTIAL
Status
• CSF, Plasma and three segments of spinal cord from normal rats
• CSF, plasma and three segments of spinal cord 5h after injury
• GC-MS and solid phase NMR
AcurePharma 40/77CONFIDENTIAL
GC/MS PROFILE
Rawdata
0
50000000
100000000
150000000
200000000
2500000001
19
37
55
73
91
10
9
12
7
14
5
16
3
18
1
19
9
21
7
23
5
25
3
Serie1
Serie2
AcurePharma 41/77CONFIDENTIAL
-10
-5
0
5
10
-10 0 10
t[2
]
t[1]
CSF_070130_Reg_process.M3 (PLS-DA), Untitledt[Comp. 1]/t[Comp. 2]
Colored according to classes in M3
Class 1Class 2
AcurePharma 42/77CONFIDENTIAL
-0,20
-0,10
0,00
0,10
-0,120 -0,100 -0,080 -0,060 -0,040 -0,020 0,000 0,020 0,040 0,060 0,080 0,100 0,120 0,140
w*c
[2]
w*c[1]
CSF_070130_Reg_process.M3 (PLS-DA), Untitledw*c[Comp. 1]/w*c[Comp. 2]
XY
Win001_C01Win001_C02
Win002_C01
Win002_C02
Win003_C01
Win003_C02
Win003_C03
Win004_C01
Win004_C02
Win005_C01Win005_C02
Win005_C03
Win005_C04
Win005_C05
Win006_C01
Win006_C02
Win006_C03Win007_C01
Win008_C01
Win008_C02
Win008_C03
Win008_C04
Win008_C05Win009_C01
Win009_C02
Win010_C01
Win010_C02
Win011_C01
Win011_C02
Win012_C01Win012_C02
Win012_C03Win012_C04
Win012_C05
Win012_C06
Win013_C01Win013_C02
Win013_C03
Win013_C04Win014_C01
Win015_C01
Win015_C02Win015_C03Win015_C04
Win015_C05
Win015_C06
Win015_C07
Win016_C01Win016_C02
Win016_C03
Win016_C04Win016_C05Win016_C06
Win016_C07
Win017_C01
Win017_C02
Win017_C03
Win017_C04
Win017_C05
Win017_C06
Win017_C07
Win018_C01
Win018_C02 Win019_C01
Win019_C02
Win019_C03
Win019_C04
Win020_C01Win020_C02Win020_C03
Win020_C04
Win021_C01
Win021_C02
Win021_C03Win021_C04
Win021_C05Win021_C06Win022_C01
Win022_C02Win023_C01
Win023_C02Win023_C03
Win023_C04Win023_C05
Win024_C01Win024_C02
Win024_C03
Win024_C04
Win024_C05
Win024_C06
Win024_C07
Win024_C08
Win024_C09
Win024_C10
Win024_C11
Win024_C12
Win024_C13
Win024_C14
Win025_C01
Win025_C02
Win026_C01Win026_C02
Win026_C03
Win027_C01
Win027_C02
Win027_C03Win027_C04
Win028_C01
Win028_C02
Win028_C03
Win028_C04
Win028_C05
Win028_C06
Win028_C07
Win028_C08
Win028_C09
Win029_C01
Win029_C02
Win029_C03 Win029_C04Win029_C05
Win029_C06
Win030_C01
Win030_C02
Win031_C01
Win031_C02
Win031_C03
Win031_C04
Win031_C05
Win031_C06
Win032_C01
Win032_C02
Win032_C03
Win032_C04
Win033_C01
Win033_C02
Win033_C03
Win033_C04
Win033_C05
Win033_C06
Win033_C07Win033_C08
Win034_C01
Win034_C02
Win034_C03
Win034_C04
Win034_C05Win035_C01
Win036_C01Win036_C02
Win036_C03Win036_C04
Win037_C01Win037_C02
Win037_C03
Win037_C04
Win037_C05
Win037_C06
Win037_C07
Win038_C01
Win038_C02Win038_C03 Win038_C04
Win038_C05
Win039_C01
Win039_C02
Win039_C03
Win039_C04
Win040_C01
Win040_C02
Win040_C03
Win041_C01
Win041_C02
Win041_C03Win041_C04
Win042_C01
Win042_C02Win042_C03
Win042_C04
Win042_C05Win043_C01
Win043_C02Win043_C03
Win043_C04
Win044_C01
Win044_C02
Win044_C03
Win044_C04Win045_C01 Win045_C02
Win045_C03
Win045_C04
Win045_C05
Win046_C01
Win046_C02
Win046_C03
Win046_C04
Win046_C05
Win047_C01Win047_C02
Win047_C03Win047_C04
Win048_C01
Win048_C02
Win049_C01Win049_C02
Win050_C01
Win050_C02
Win051_C01
Win051_C02
Win051_C03
Win052_C01
Win052_C02
Win053_C01
Win053_C02
Win054_C01
Win054_C02
Win054_C03
Win055_C01
Win055_C02
Win055_C03
Win055_C04
Win056_C01Win056_C02
Win056_C03
Win056_C04
Win057_C01
Win057_C02
Win058_C01
Win058_C02
Win059_C01
Win060_C01
Win061_C01
Win061_C02
Win061_C03
Win061_C04
Win061_C05
Win062_C01
Win062_C02
Win062_C03
Win063_C01
Win063_C02
Win064_C01
Win064_C02
Win064_C03Win065_C01
Win066_C01
Win066_C02
Win066_C03
Win067_C01
Win067_C02
Win067_C03
Win068_C01
Win068_C02
$M3.DA1
$M3.DA2
AcurePharma 43/77CONFIDENTIAL
- 2
- 1
0
1
2
- 5 - 4 - 3 - 2 - 1 0 1 2 3 4 5
t[2]
t[1]
CSF_070130_Reg_process.M6 (PLS-DA), Untitled
t[Comp. 1]/t[Comp. 2]Colored according to classes in M6
Class 1
Class 2
AcurePharma 44/77CONFIDENTIAL
-0,40
-0,20
0,00
0,20
0,40
0,60
-0,50 -0,40 -0,30 -0,20 -0,10 0,00 0,10 0,20 0,30 0,40 0,50
w*c
[2]
w*c[1]
CSF_070130_Reg_process.M6 (PLS-DA), Untitledw*c[Comp. 1]/w*c[Comp. 2]
XY
Win011_C01
Win017_C02
Win024_C07
Win030_C02
Win033_C01
Win037_C05
Win055_C01
Win057_C02
Win067_C01$M6.DA1
$M6.DA2
AcurePharma 45/77CONFIDENTIAL
Arthritis – Diagnosis
• Early diagnosis critical
� Reduce symptoms of wear
� More successful treatment with early medication*
• Rheumatoid arthritis
� Physical examination, antibodies (today not specific for RA),X-ray
• Osteoarthritis
� Physical examination, X-ray
• New diagnostic tools are needed…
* http://www.reumatikerforbundet.org/start.asp?sida=3955
AcurePharma 46/77CONFIDENTIAL
Case study 1: Sample information
Deconvolution (MCR v.1.13)
� 59 Samples
� 181 Spectral Profiles
Outliers (possible)
� Not primary detected
Included Isotope labeled mixture� 11 Internal standards + Methyl
Stearate
Main Groups
�3 Health Groups•• ControlControl
•• RaRa
•• OaOa
Analysed by gaschromatography/time of flightmass spectrometry (GC/TOFMS)
AcurePharma 47/77CONFIDENTIAL 47CONFIDENTIAL
Rheumatoid Arthritis – brief background
• Worldwide prevalence of approximately 1%*
• Autoimmune disease, the body attacks itself, aetiology largely unknown**
• Treatment; irreversible disease, no known cure, medication to maintain mobility and ease pain
• Early diagnosis critical
� Reduce symptoms of wear
� More successful treatment with early medication*
• Diagnosis for rheumatoid arthritis
� Physical examination, antibodies (today not specific for RA),X-ray, MRI
• New diagnostic tools are needed…
* Feldmann, M. et al., Cell. 85: 307-310 (1996)** Krishnan, E., Joint Bone Spine. 70: 496-502 (2003)
AcurePharma 48/77CONFIDENTIAL
Isotope Labeled Internal Standards
L-Proline-13C5-TMS (221 sec) Win04_C02
Succinic Acid-D4-2TMS (225 sec) Win04_C07
Salicylic Acid-D6-2TMS (265 sec) Win08_C01
Di-Na-alfa-Ketoglutarat-13C4-2TMS (277 sec) Win09_C03
L-Glumatic Acid-13C5-15N (285 sec) Win10_C02
Putrescine-D4-4TMS (305 sec) Win11_C07
Myristic Acid-13C3-TMS (324 sec) Win13_C03
D-Glucose-13C6-5TMS (331 sec) Win14_C01
Hexadecanoic Acid-13C4-TMS (354 sec) Win16_C02
Methyl stearate (365 sec) Win17_C03
Sucrose-13C12-8TMS (430 sec) Win24_C01
Cholesterol-D7-TMS (483 sec) Win29_C02
AcurePharma 49/77CONFIDENTIAL
2008-06-11 Johan Trygg, Research Group forChemometrics, Umeå University
49
Principal Component Analysis
AcurePharma 50/77CONFIDENTIAL
Data Overview
-10
-9
-8
-7
-6
-5
-4
-3
-2
-1
0
1
2
3
4
5
6
7
8
9
10
-14 -13 -12 -11 -10 -9 -8 -7 -6 -5 -4 -3 -2 -1 0 1 2 3 4 5 6 7 8 9 10 11 12 13 14
t[2]
t[1]
MVA_A_norm-IS_HS.M1 (PCA-X), All 3 groups, (28 excl)
R2X[1] = 0,149112 R2X[2] = 0,0715346
Ellipse: Hotelling T2 (0,95)
CCC
C
C
CC
C
C
C
C
CC
C
C
CC
C
C
Ra
Ra
Ra
Ra
Ra
Ra
RaRa
Ra
Ra
Ra
RaRaRa
Ra
RaRa
Ra
Ra
Oa
Oa
Oa
Oa
Oa
Oa
OaOa
Oa
Oa
Oa
Oa
Oa
Oa
Oa
Oa
Oa
Oa
Oa
Oa
SIMCA-P 11 - 2006-04-10 13:41:52
-9
-8
-7
-6
-5
-4
-3
-2
-1
0
1
2
3
4
5
6
7
8
9
-14 -13 -12 -11 -10 -9 -8 -7 -6 -5 -4 -3 -2 -1 0 1 2 3 4 5 6 7 8 9 10 11 12 13 14
t[2]
t[1]
MVA_A_norm-IS_HS.M2 (PLS-DA), All 3 groups, (28 excl)
R2X[1] = 0,148724 R2X[2] = 0,0635685
Ellipse: Hotelling T2 (0,95)
C
C
C
CCC
C
CC
C
C
C
C
C
CCCC
C
Ra
Ra RaRa
Ra
RaRa
Ra
Ra
Ra
Ra
Ra
Ra
Ra
Ra
Ra
Ra
RaRa
Oa OaOaOa
Oa
Oa
OaOa
Oa
Oa
Oa
Oa
Oa
Oa
OaOa
Oa
Oa
Oa
Oa
SIMCA-P 11 - 2006-04-10 13:42:46
PCA – 181 variables PLS-DA – 181 variables
AcurePharma 51/77CONFIDENTIAL
2008-06-11 51
The OPLS method
PLS, MLR, PCR, RR etc...
Mixes Y-orthogonal and Y-predictive variationUni-directional, Models Y from X
Models X AND Y Separates Orthogonal and Predictive variation-e.g. ‘between block’ from ‘within block’
Bi-directional Only uses predictive variation for modeling Y
OPLSOrthogonal Projections of Latent Structures
AcurePharma 52/77CONFIDENTIAL 52CONFIDENTIAL
Rheumatoid arthritis: Control vs. RAOPLS-DA*- 204 putative biomarkers
*Bylesjö, M.; Rantalainen, M.; Cloarec, O.; Nicholson, J. K.; Holmes, E.; Trygg, J.,
OPLS discriminant analysis: combining the strengths of PLS-DA and SIMCA classification.
J. Chemometrics, 20, 341-351 (2006).
-8
-6
-4
-2
0
2
4
6
8
-10 0 10
t[2]O
t[1]P
(OPLS), opls-da, all, uv scaled
R2X[1] = 0.16 R2X[2] = 0.07
Ellipse: Hotelling T2 (0.95)
ControlRA
SIMCA-P 11 - 29/02/2008 09:19:58
Within group dynamics
not important for class separation
Group separating direction
Specific metabolites for healthy and diseased
AcurePharma 53/77CONFIDENTIAL 53CONFIDENTIAL
Rheumatoid arthritis: Control vs. RA
• Significant (subset) metabolites for separation of RA samples from healthy controls.� Variables represents endogenous metabolites
� Data set includes identified and yet to be identified potential biomarkers for diagnosis of RA
- 0 , 4
- 0 , 3
- 0 , 2
- 0 , 1
- 0 , 0
0 , 1
0 , 2
0 , 3
Win
02
1_C
02
Win
01
9_C
01
Win
02
0_C
02
Win
01
0_C
06
Win
01
8_C
01
Win
02
2_C
04
Win
02
9_C
03
Win
02
0_C
05
Win
01
2_C
11
Win
00
4_C
08
Win
02
9_C
04
Win
03
3_C
02
Win
03
2_C
01
Win
00
6_C
05
Win
01
7_C
01
Win
00
1_C
07
Win
00
1_C
06
Win
03
5_C
02
Win
01
9_C
06
Win
02
6_C
01
Win
00
8_C
02
Win
02
3_C
02
Win
01
3_C
02
Win
00
4_C
06
Win
01
2_C
03
Win
02
8_C
02
Win
02
8_C
01
Win
01
0_C
03
Win
02
2_C
01
p[1
]
V a r I D ( P r i m a r y )
M V A _ A _ n o r m - I S _ H S . M 6 ( P C A - X ) , R a v s . O a , 9 9 % C I 2 9 v a r p [ C o m p . 1 ]
R 2 X [ 1 ] = 0 , 3 6 9 8 9 3 S I M C A - P 1 1 - 2 0 0 6 - 0 4 - 1 0 1 4 : 1 2 : 4 0
Positive correlation to RaPositive correlation to Ra
Positive correlation to Control
AcurePharma 54/77CONFIDENTIAL 54CONFIDENTIAL
RA: Comparison of the human case and animal models
• Great overlap of metabolites between humans and animals
� Different metabolites show overlap in different animal models
� Allows for identification of relevant animal models
� Selection of model system for treatment studies
BMHuman Rheumatoid
Arthritis
Mouse Collagen
Induced Arthritis
Rat Adjuvant
Induced Arthritis
EC001 ↑ na Na
EC002 ↑ ? ?
EC003 ↑ ↓ ↓
EC004 ↑ 0/ ↓ ↓
EC005 ↓ na na
EC006 ↓ ↓ ↓
EC007 ↓ ↓ ↓
EC008 ↓ ↓ ↑
EC009 ↓ ↓ ↓
EC010 ↓ ↑ ↑
EC011 ↓ 0/ ↓ ↓
EC012 ↓ na na
EC013 ↓ ↓ ↓
EC014 ↓ ↓ ?
EC015 ↓ ↓ ↓
EC016 ↓ ? ↓
EC017 0 ↓ ↓
EC018 ↑ ↑ ↓ / ?
EC019 ↓ ↓ ↓
EC020 ↓ ↓ / ? ↓
EC021 ? ↑ / ? ↑
EC022 ↓ ↓ ↓
EC023 0 ↓ ↓
EC024 ↑ ↓ 0/ ↑
EC025 ↑ ↓ ↓
EC026 0/ ↑ ↓ ↑
AcurePharma 55/77CONFIDENTIAL 55CONFIDENTIAL
RA: Comparison of therapies in animal model
• Metabolites levels are affected by administered therapeutics
� New drug (X) restore levels in more metabolites compared to MTX*
� Useful in development of novel drugs
� Tool in clinical studies to verify therapeutic effect in clinical studies
� Concomitant development of novel drug and diagnostic test, theranostics?
Vehicle MTXX
1mg
X
3mg
X
10mg
EC004 0/ ↑ ↓ ↓ 0/ ↓ ↓
EC006 0/ ↑ / ? 0/ ? 0 ↑ ↑
EC007 ↓ 0/ ↑ 0/ ↑ 0/ ↓ ↑
EC009 0 ↑ ↓ ↑ ↑
EC010 ↑ ↑ ↑ ↑ ↑
EC011 0 0/ ↓ ↓ 0/ ↓ ↑
EC012 0/ ↓ ↑ 0/ ↓ ↑ ↑
EC013 ↑ 0/ ↑ 0/ ↓ ↑ 0/ ↑
EC014 ↑ 0/ ? ↑ ↑ ↑
EC015 0/ ↑ ↑ 0/ ↓ ↑ ↑
EC016 0 ↓ ↑ ↑ ↓
EC017 ↓ ↓ ↓ ↓ ↓
EC018 ↓ ↓ ↓ 0/ ↑ 0/ ↑
EC019 ↓ ↓ ↓ ↓ ↓
EC022 ↑ ↑ 0/ ↑ ↑ ↑
EC023 ↓ 0/ ↓ ↓ ↓ ↓
EC024 ↓ ↓ ↓ ↓ ↓
EC025 ↓ ↓ ↓ ↓ ↓
EC026 ↑ ↑ ↑ ↑ ↑
*MTX, methotrexate
AcurePharma 56/77CONFIDENTIAL
Theranostics• Theranostic approaches aim to target therapies to
those patients most likely to benefit by combining with diagnostic or prognostic biomarkers
• Greatest benefit is realised with therapies that are effective, but in unknown patient sub-populations –HA therapy in OA is an excellent example
• An effective theranostic can drive sales of therapies as illustrated
• Regulation can be achieved within current pivotal clinical trial protocols
• Potential for the approach to track therapeutic efficacy
Benefits• Identification of responder/non responders populations
• Targeted patient recruitment to pivotal clinical trials, resulting in shorter, lower risk studies
• Stronger claims through mechanism of action knowledge and potential for data on therapeutic efficacy
• Improved HE evidence and addressing payer “entitlement” concern will increase payer uptake
• Much higher adoption in smaller market size – 50% penetration of 50% market size is better than 1-5% of the
full market
Reduced timeto peak sales
Enhancedpeak sales
ExtendedLife cycle
0 5 years 10 years
Sale
s
Faster to
market
+ theranostic
- theranostic
Adapted from Gilham (2002) Theranostics: an emerging tool in drug discovery and commercialisation. Drug Discovery
World
Reduced timeto peak sales
Enhancedpeak sales
ExtendedLife cycle
0 5 years 10 years
Sale
s
Faster to
market
+ theranostic
- theranostic
Adapted from Gilham (2002) Theranostics: an emerging tool in drug discovery and commercialisation. Drug Discovery
World
AcurePharma 57/77CONFIDENTIALCONFIDENTIAL
Case study 2: Functional foods study• NMR spectroscopy to detect metabolism changes due to food
supplement
• N = 47 biofluid samples
� 9 healthy individuals
� Given prepared foodstuff for consumption
� Multiple visits – document effect over time
• K = 32 768 variables
� NMR shifts
• Goal: Effect of food supplement?
� If any, what metabolites?
AcurePharma 58/77CONFIDENTIAL
2008-06-11 Johan Trygg, Research Group forChemometrics, Umeå University
58
Chemometrics- the information aspect for studying
complex systems1. Define the aim
� What do we want?2. Selection of objects (e.g. samples, time points, internode,
experiments)
• Design of Experiments (DOE)
• Multivariate design (MVD)3. Sample preparation and characterisation
• Experimental protocol (e.g. FTIR, GCMS,Microarray)
• Data processing (e.g. normalisation of FTIR, microarrays)
4. Evaluation/Validation of collected data
• Exploratory analysis, Interpretation &Visualization
• Dynamic study
AcurePharma 59/77CONFIDENTIAL
2008-06-11 Johan Trygg, Research Group forChemometrics, Umeå University
59
• Functional foods: Foodstuffs with a documented health-promoting effect – besides energy addition
• Centre for Human Studies of Foodstuffs, Sweden
� Inclusion/exclusion criteria
� 9 individuals given prepared foodstuff
� Multiple visits – document effect over time
Example:Functional foods study
AcurePharma 60/77CONFIDENTIAL
Case study 2: Overview of allindividuals/samples
Clear separation.
But �… due to different sampling periods
AcurePharma 61/77CONFIDENTIAL
2008-06-11 61
Prediction of triglyceridesRaw NMR data vs Metabolic effect
modelling
Prediction of triglycerides
Metabolic effect modelling
Prediction of triglycerides
Raw NMR data (major bias)
AcurePharma 62/77CONFIDENTIAL
Case study 2: Modelling dynamic metabolomic time series data
• Assumption 1: The metabolic baseline can be different over all individuals
• Assumption 2: The metabolic effect of the treatment is similar over all individuals
AcurePharma 63/77CONFIDENTIAL
Case study 2: Overview of all
individuals/samples… after pre-processing/filtering
AcurePharma 64/77CONFIDENTIAL
Case study 2: Overview of all
individuals/samples… after pre-processing/filtering
AcurePharma 65/77CONFIDENTIALCONFIDENTIAL
Case study 2: Overview of all
individuals/samples… after pre-processing/filtering
Möte Q-MED 2007-11-08
AcurePharma 66/77CONFIDENTIALCONFIDENTIAL
Case study 2: Overview of all
individuals/samples… after pre-processing/filtering
Möte Q-MED 2007-11-08
AcurePharma 67/77CONFIDENTIAL
Case study 2: Metabolic profiles for each individual
Assumption: the effect of foodstuff treatment
CONFIDENTIALMöte Q-MED 2007-11-08
AcurePharma 68/77CONFIDENTIALCONFIDENTIAL
Case study 2: Prediction of effect in NMR profile
Model
Prediction
AcurePharma 69/77CONFIDENTIALCONFIDENTIAL
Case study 2: Prediction of health effect inendogenous metabolites (”biomarkers”)
Model
Prediction
Health effect shown, as metabolite concentration
decreases after consumption of foodstuff
AcurePharma 70/77CONFIDENTIAL
Case study 2: Analysis of health effect in
endogenous metabolites ”biomarkers”)
AcurePharma 71/77CONFIDENTIAL
Case study 2: Conclusions
• Effect of food supplement established
� Increase in Myo-Inositol
� Decrease in triglycerides
• Opens new possibilities for development of functional foods
� Proof of physiological health effect
� Target identification
AcurePharma 72/77CONFIDENTIAL
Define the aim/aims
Define the aim/aims
Design of experiments
Design of experiments
Training setTest set
Training setTest set
Multivariatedesign
Multivariatedesign
General analysis strategy
Sampleacquisition
Sampleacquisition
SpectroscopyProcessing
SpectroscopyProcessing
Datapre-processing
Datapre-processing
Multivariate data analysis
Multivariate data analysis
PCAPCA
PCA-SIMCAPCA-SIMCA
PLS-DAPLS-DA
OPLSOPLS
SamplePre-treatment
SamplePre-treatment
DESIGN
AcurePharma 73/77CONFIDENTIAL
Strategy1. Formalize the aim
• What do we want?
2. Selection of objects• Design of Experiments (DOE)• Multivariate design (MVD)
3. Sample preparation and profiling of human and animal samples• In vivo, in vitro samples• Blood, Urine, Cerebral Spinal Fluid (CSF)• Synovial fluid (joint), Bowel fluids, Feces, Tissues
4. Integration and evaluation of collected data • Exploratory analysis, Interpretation & Visualization• Prediction models • Patterns• Target identification
5. Identify and define IPR opportunities and strategies
AcurePharma 74/77CONFIDENTIAL
Acureo m i c sAcureo m i c s
Osteoarthritis (patent application filed)
Rheumatoid arthritis (patent application filed)
Fibromyalgia
Chronic Fatigue Syndrome
Type 1 Diabetes (patent application filed)
Projects in disease diagnostics Services CRO
Biofluid profiling
Biomarker identification
Toxicity monitoring
AcurePharma 75/77CONFIDENTIAL
Summary
• Historical data
� Always a good starting point for analysis (PCA)
� Determine data structure, preferred format/output
� Get acquainted with the process and the data
� Find “hidden” information
� Good starting point for discussions
• Determine� The aim – is there a defined stop criteria
(yield, purity, accepted batches, ID important/”sensitive”variables etc.)
� Important to define prior to investigation � know when to stop
� The experimental domain(variables, settings, responses)
AcurePharma 76/77CONFIDENTIAL
Summary
• Design of Experiments (DoE)
� Simplify analysis
� Ensure a systematic variation in the investigated experimental domain
� Small design within the defined limits
� (Design in historical data)
• Analysis
� PCA (SIMCA etc.)
� PLS, PLS-DA
• Next step
� Aim(s) reached
� Important variables
� New optimal experiments
AcurePharma 77/77CONFIDENTIAL
Acknowledgements
• Per Lek, AcurePharma
• Thomas Moritz, UPSC
• Johan Trygg, Umeå University� Rasmus Madsen, Umeå University
• Elisabeth Seifert, AcurePharma
• Jon Gabrielsson, AcureOmics
• Johan Olsson, Uppsala University
• Mattias Hedenström, Umeå University
• Katrin Lundstedt-Enkel, Uppsala University
