Overview 1 March 2011 1 September 2012 On behalf of the DDMoRe - - PowerPoint PPT Presentation

▶
overview
SMART_READER_LITE
LIVE PREVIEW

Overview 1 March 2011 1 September 2012 On behalf of the DDMoRe - - PowerPoint PPT Presentation

Jan 02, 2024 231 likes •541 views

Overview 1 March 2011 1 September 2012 On behalf of the DDMoRe consortium The Productivity Gap in Pharma R&D 60 $55 50 50 45 40 New Drug Approvals 40 Pharma R&D ($ billions) 35 30 30 25 20 20 15 10 10 5 0 0 92 92

slide-1
SLIDE 1

On behalf of the DDMoRe consortium

Overview

1 March 2011 – 1 September 2012

slide-2
SLIDE 2

WCoP-Seoul-2012

2

Source: Burrill & Company; US Food and Drug Administration.

10 20 30 40 50 60 5 10 15 20 25 30 35 40 45 50 $55

New Drug Approvals

Pharma R&D ($ billions)

92 92 93 93 94 94 95 95 96 96 97 97 98 98 99 99 00 00 01 01 02 02 03 03 04 04 05 05 06 06 07 07

The Productivity Gap in Pharma R&D

2

slide-3
SLIDE 3

WCoP-Seoul-2012

3

Innovative Medicines Initiative: the Largest PPP in Life Sciences R&D

  • Key concepts
  • Open innovation
  • Pre-competitive research

2 Billion Euro

1 Billion € 1 Billion €

Public Private Partnership

The Four Pillars of the Innovative Medicines Initiative

3

slide-4
SLIDE 4

WCoP-Seoul-2012

5

Participants

are a unique combination of model builders, model users, software developers and teachers

5

slide-5
SLIDE 5

WCoP-Seoul-2012

6

Participants

are a unique combination of model builders, model users, software developers and teachers

6

slide-6
SLIDE 6

WCoP-Seoul-2012

8

DDMoRe – The Vision

Major deliverables

  • Data contains raw information,

which is difficult to share

  • IP, CDISC
  • Models
  • represent an interpretation,

understanding of the data (given experimental conditions)

  • allow to predict the future

with uncertainty

  • are an intellectual container
  • f the knowledge

Modelling Library

Shared knowledge

8

slide-7
SLIDE 7

WCoP-Seoul-2012

9

Modelling Library

Shared knowledge

Model Definition Language System interchange standards

Standards for describing models, data and designs … has standards at the core of its strategy …

DDMoRe – The Vision

Major deliverables

9

slide-8
SLIDE 8

WCoP-Seoul-2012

10

DDMoRe – The Vision

Major deliverables

Modelling Library

Shared knowledge

Modelling Framework

A modular platform for integrating and reusing models; shortening timelines by removing barriers

Model Definition Language System interchange standards

Standards for describing models, data and designs … but the framework will put the system into life …

10

slide-9
SLIDE 9

WCoP-Seoul-2012

11

DDMoRe – The Vision

Major deliverables

Modelling Library

Shared knowledge

Modelling Framework

A modular platform for integrating and reusing models; shortening timelines by removing barriers

Model Definition Language System interchange standards

Specific disease models

Examples from high priority areas

Standards for describing models, data and designs … PoC: implementation and evolution of DA models

11

slide-10
SLIDE 10

WCoP-Seoul-2012

12

Modelling Library

Shared knowledge

Modelling Framework

A modular platform for integrating and reusing models; shortening timelines by removing barriers

Model Definition Language System interchange standards

Specific disease models

Examples from high priority areas

Standards for describing models, data and designs

DDMoRe – The Vision

Major deliverables

12

slide-11
SLIDE 11

WCoP-Seoul-2012

15

Proliferative cells Circulating WBC Feedback= WBC(t) WBC0 

Non-mitotic cells Mean Transit Time (MTT) = 4/ktr

Drug concentration

Slope

(or Emax-model)

kprol=ktr ktr ktr ktr ktr ktr D2-receptor antagonists +

Dopamine Prolactin Blood

  • +

1 - kDA kDA Kin kout

Drug-specific parameter: Ki

Agonist-antagonist interaction model

kdeath

Conc

ke (=0) kgrowth

S R

kdeath kSR kdrug

= Emax·Ce/(Ce+Ec50

)

Ce

ke0 ke0

S=sensitive, proliferating R=resting, insensitive

S+R+

Delayed effect

Drug Bacteria

Rheumatoid Arthritis – ACR20 Lacroix et al., CPT 2009 Oncology – Myelosuppression Friberg et al., JCO, 2002 Schizophrenia – Prolactin elevation Friberg et al., CPT 2009 Bacteria kill of antibiotics Nielsen et al., AAC 2007 Glucose - Insulin Silber et al., JCP 2007 Tumor growth – Xenografts Simeoni et al., 2004

1 Responder Non- Responder 2 Dropped

  • ut

Pr00 of remaining non-responder Pr11 of remaining responder Pr10 of becoming responder Pr01 of becoming non-responder Pr12 of dropping out Pr02 of dropping out

Examples of models to be implemented

15

slide-12
SLIDE 12

WCoP-Seoul-2012

Model Exchange

19

NONMEM MONOLIX WinBUGS R MatLab PK/PD Model in MML

Model Repository MDL

19

slide-13
SLIDE 13

WCoP-Seoul-2012

Outcome

MML specification DDMoRe ML SBML SED-ML NuML CellML PharML libMML (WP 2.3: API)

23

slide-14
SLIDE 14

WCoP-Seoul-2012

Developing the Specification

MSSMml definition build prototype implementation Code generator/tr anslator xml test cases Executable Model refine definition tests work? Yes No expand definition

24

slide-15
SLIDE 15

WCoP-Seoul-2012

Model Exchange

29

NONMEM MONOLIX WinBUGS R MatLab PK/PD Model in MML

Model Repository MDL

29

slide-16
SLIDE 16

WCoP-Seoul-2012

Objects Sub-Language Language

MDL MCL Data Parameter Model Task TEL R Command TEL Command

  • User defines new models using MCL
  • TEL provides script-based updating of MCL objects and execution settings,

retrieval of task output

31

MDL languages and objects

31

slide-17
SLIDE 17

WCoP-Seoul-2012

Data

HEADER FILE INLINE RSCRIPT DESIGN

Model

INPUT VARIABLES STRUCTURAL PARAMETERS VARIABILITY PARAMETERS GROUP VARIABLES RANDOM VARIABLE DEFINIITON INDIVIDUAL VARIABLES LIBRARY ODE MODEL PREDICTION ESTIMATION SIMULATION OUTPUT VARIABLES

Parameter

STRUCTURAL VARIABILITY

Task

DATA PARAMETERS ESTIMATE SIMULATE

  • Block structure separates structural

(fixed) and variability (random) parts of a model

  • Modularity and use of sub-component

models.

32

MCL objects and blocks

32

slide-18
SLIDE 18

WCoP-Seoul-2012

MCL – Examples

NM-TRAN

$INPUT ID TIME AMT ODV INSU TOTG CMT BW EVID RATE DV TYPE OCC ;data contains dosing records, glucose(1), hot glucose(3) and insulin(2) $DATA data_OGTT.csv IGNORE=@

MCL Data Object

OGTT_IGI_dat = dataobj{ HEADER ID=list(type=categorical) TIME=list(type=continuous, units="h") CMT=list(type=categorical) ... BW=list(use=continuous, units="kg") ... OCC=list(type=categorical) # end HEADER FILE source="data_OGTT.csv" unjumble="NONMEM" ignore_char="@" # end FILE ) # end data object0

34 34

slide-19
SLIDE 19

WCoP-Seoul-2012

TEL Object Structure Principles

35

TEL

Task Command (Target Application) R Command (R Environment)

Task object defines functions

Uses R language syntax May call target applications such as NONMEM, Monolix, etc. through MCL task objects or pass commands to R environment

35

slide-20
SLIDE 20

WCoP-Seoul-2012 45

slide-21
SLIDE 21

WCoP-Seoul-2012

Interoperability Framework

50

slide-22
SLIDE 22

WCoP-Seoul-2012

53

Interoperability Framework Data Flow

53

slide-23
SLIDE 23

WCoP-Seoul-2012

  • n behalf of the DDMoRe Consortium

58

Model Repository in DDMoRe

58

slide-24
SLIDE 24

WCoP-Seoul-2012

  • n behalf of the DDMoRe Consortium

59

Functionalities (overview)

59

slide-25
SLIDE 25

WCoP-Seoul-2012

$DESCRIPTION PKPD model $PSI ka V Cl Imax IC50 Rin kout $EQUATION k=Cl/V E_0 = Rin/kout Cc = Ac/V DDT_Ad = -ka*Ad DDT_Ac = ka*Ad - k*Ac DDT_E = Rin*(1-Imax*Cc/(Cc+IC50)) - kout*E

64

A prototype of the CTS

64

slide-26
SLIDE 26

WCoP-Seoul-2012

Task 6.2 Adaptive Optimal Design

  • A survey among EFPIA participants has identified the level of current usage of
  • ptimal design software and the expectations of the industry for future

capabilities.

  • Problems that will be studied:
  • Model robustness (model averaging/ selection)
  • Local versus Robust Optimal design (pros & cons)
  • How to include the previous information in the design calculations
  • Estimation
  • Effect of early cohorts as the driving force in an adaptive design
  • What to optimize?
  • How to optimize
  • Within subject adaptations vs. Between subject adaptations

65

65

slide-27
SLIDE 27

WCoP-Seoul-2012

Task 6.3 Diagnostic tools

New diagnostic tools are necessary for model selection and model assessment. Specific areas of interest are:

  • New diagnostics for repeated time-to-event models
  • Optimal covariate model building strategies
  • Use of simulation techniques to assess various model diagnostics (VPC, NPC and NPDE
  • f model diagnostics)
  • Fit output from NLME to simpler models to help in diagnosing and building
  • From visual diagnostic tools to inference and decision tools
  • Sampling from conditional distributions for model assessment

66

66

slide-28
SLIDE 28

WCoP-Seoul-2012

Training & Education (T&E)

F2F course curri- culum/ material

F2F Diabetes

Survey => requi- rements for DD M&S Web- based self-edu- cational training Mecha- nisms/ Process of intern- ship

F2F Onco, Beg. F2F Onco, Adv. F2F Oth., Beg. F2F Oth., Adv. F2F Infect. F2F Safety

Objectives:

  • To create a landscape of technical and conceptual requirements in Drug/Disease

Modelling and Simulation (DD M&S)

  • To develop a Training and Education program incl. material in DD M&S
  • n PhD and postdoc level

69

69

slide-29
SLIDE 29

WCoP-Seoul-2012

Website and Newsletter

www.ddmore.eu

73