Assay Design and Quality Control in

HTS and Lead Optimisation.

• Criteria for assay development and acceptance.

• Factors affecting assay variability.

• Statistical analysis in assay development and screening.

• So what?

“To knowledge by measurement”

Kammerlingh

“Science is Measurement”

Lord Kelvin

“An assay is only worth doing if it generates reliable SAR. “

HTS Technologies

Scintillation Proximity Assay

Homogenous Time Resolved Fluorescence

Fluorescence Polarisation

Fluorescence Correlation Spectroscopy

Fluorescence Intensity Distribution Assay

Alphascreen

Assay Acceptance Criteria

Reproducible.

Predictive

Physical Causes of Assay Variation

• Liquid Handling Errors

• Reagent adhesion

• Evaporation

• Mixing

• Incubation conditions

• Human Error

• Biochemical Assays:

• pH• Temperature• Ion concentration• Reagent Stability• Reagent Aggregation• Reagent Solubility• Order of reagent addition• Solvent effects

Biological Causes of Assay Variation

2. Cell Based Assays As for Biochemical assays, plus:

• Cell culture plastics• Culture media• Serum• Cell cycle• Passage number• Solvent effects• Infection

Biological Causes of Assay Variation

Optimisation of Assay Parameters

• Hi tech: Automated Assay Development. JMP Statistica Design Expert

• Low Tech: Trial and Error

Statistical Analysis

0

1000

2000

3000

4000

5000

6000

7000

A1 A4 A7A10B1 B4 B7

B10C1 C4 C7C10 D1 D4 D7

D10 E1 E4 E7E10 F1 F4 F7

F10G1 G4 G7G10H1 H4 H7

H10

Sample Number

cpm

Statistical Analysis

SeparationBand

Data variabilityBand

Data variabilityBand

s c

Fre

quen

cy

Assay Signal

3s 3c

Statistical Analysis

Statistical Analysis: Z Factor

3SD of sample + 3SD of control mean of sample – mean of control

Z = 1 –

Zhang, Chung and Oldenburgh, 1999

Z Factor Assay Screening

1 SD = 0 or dynamic range Ideal assay

1 > Z >= 0.5 Separation band is large Excellent assay

0.5 > Z > 0 Separation band is small Double assay

0 No separation band, sample signal Yes / No assay and control signal variation touch

< 0 No separation band, sample signal Unusable assay and control signal variation overlap

Statistical Analysis: Z Factor

SeparationBand

Data variabilityBand

Data variabilityBand

s c

Fre

quen

cy

Assay Signal

3s 3c

Z = 1 SD = 0 or Dynamic range Ideal1> Z >= 0.5 Separation band is large Excellent

Statistical Analysis

Data variabilityBand

Data variabilityBand

s c

Fre

quen

cy

Assay Signal

3s 3c

0.5 > Z > 0 Separation band is small Double assayZ = 0 No separation band Yes / No assay

Statistical Analysis

s c

Fre

quen

cy

Assay Signal

Statistical Analysis

3s 3c

Z < 0 No separation band Unusable assay

Quality Control of Screening Data

• Standards

Inter Assay Variability

• Appropriate plate layouts

Plate DriftIntra Assay Variability

1 2 3 4 5 6 7 8 9 10 11 12A NSB TB C1 C2 C3 C4 C5 C6 C7 C8 C9 C10B NSB TB C11 C12 C13 C14 C15 C16 C17 C18 C19 C20C NSB TB C21 C22 C23 C24 C25 C26 C27 C28 C29 C30D NSB TB C31 C32 C33 C34 C35 C36 C37 C38 C39 C40E NSB TB C41 C42 C43 C44 C45 C46 C47 C48 C49 C50F NSB TB C51 C52 C53 C54 C55 C56 C57 C58 C59 C60G NSB TB C61 C62 C63 C64 C65 C66 C67 C68 C69 C70H NSB TB C71 C72 C73 C74 C75 C76 C77 C78 C79 C80

1 2 3 4 5 6 7 8 9 10 11 12A C1 C3 C5 C7 C9 C11 NSB TB C15 C17 C19 C21B C1 C3 C5 C7 C9 C11 NSB TB C15 C17 C19 C21C C1 C3 C5 C7 C9 C11 NSB TB C15 C17 C19 C21D C1 C3 C5 C7 C9 C11 NSB TB C15 C17 C19 C21E C2 C4 C6 C8 C10 C12 C13 C14 C16 C18 C20 REFF C2 C4 C6 C8 C10 C12 C13 C14 C16 C18 C20 REFG C2 C4 C6 C8 C10 C12 C13 C14 C16 C18 C20 REFH C2 C4 C6 C8 C10 C12 C13 C14 C16 C18 C20 REF

Assay Plate Layout (Rogue’s Gallery)

1 2 3 4 5 6 7 8 9 10 11 12A TB C1 C9 C17 C25 C33 C41 C49 C57 C65 C73 REFB TB C2 C10 C18 C26 C34 C42 C50 C58 C66 C74 REFC TB C3 C11 C19 C27 C35 C43 C51 C59 C67 C75 NSBD NSB C4 C12 C20 C28 C36 C44 C52 C60 C68 C76 NSBE NSB C5 C13 C21 C29 C37 C45 C53 C61 C69 C77 NSBF NSB C6 C14 C22 C30 C38 C46 C54 C62 C70 C78 TBG REF C7 C15 C23 C31 C39 C47 C55 C63 C71 C79 TBH REF C8 C16 C24 C32 C40 C48 C56 C64 C72 C80 TB

Assay Plate Layout

Usable Assay Variability

• HTS

Low false positive rate

• Lead Optimisation

Ki inter assay variability < 3 fold(Preferably < 2 fold)