Agent-based methods for translational cancer multilevel modelling

Sylvia Nagl PhD

Cancer Systems Science & Biomedical Informatics

UCL Cancer Institute

London

Main points of the talk

• Potential of agent-based modelling

• Systems biology perspective on large cell network simulation

• A new synergy between modelling and wet biology

Hanahan and Weinberg (2000) Cell 100:57-70

The hallmarks of cancer

Systems biology and medicine

• Diseases are abnormal perturbations of biological networks - through defects in molecular mechanisms or environmental stimuli

• Therapies are the interventions needed to restore networks to their normal states

1

2

3

4

5

6

7 8

910

11

12

13

14

15

16

17

18

1920

21

22

23

24

25

26

27

28

293031

32

3334

35

36

37

38

39

40

41

42

43

4445

46474849

50

51

52

53

54

5556

57

58

59

606162

63

64

65

66

67

68

6970

71

72

73

7475

76

77

78

7980

81

82

83

84

85

86

87

88

89

9091

92

93

94

95

96

97

98

99

100

101102

103

104

105

106107108

109

110

111

112

113114115

116

117

118

119

120

121

122

123

124

125

126

127128

129

130

131

132

133

134135

136

137

138

139

140

141

142

143

144

145

146

147

148

149

150

151

152

153

154155

156

157

158

159

160

161

162

163

164

165

166

167

168

169170

171

172173

174

175

176

177

178

179180

181

182 183184

185

186

187

188

189

190

191

192

193194

195

196

197

198

199

200

201

202203

204

205

206

207

208

209

210

211

212

213

214

215

216

217

218

219

220

221

222223

224

225

226

227

228

229

230

231

232

233234

Butcher et al. (2004) Nature Biotechnology 22:1253

Modelling challenge: genome to phenotype

extended genotype

elementary phenotype

Systems biology and medicine

• Fundamental question of where function lies within a cell – distributed (networks of interacting molecules)– hierarchical

• network motifs and modules • complex network connecting modules

• A globalist view of the dynamics of (large) cell networks is therefore needed

cell and tissue levels

cell networks

molecular interactions (molecular dynamics)

E-science}

Systems biology and cancer

• Given the many components of functional modules, there are different paths to disease-inducing systems failure

• A multitude of ways to ‘solve’ the problems of achieving a survival advantage in cancer cells

• Each patient’s cancer cells evolve through an independent set of genomic lesions and selective environments - a fundamental reason for differences in survival and treatment response

Likelihood of cancer cell death in response to DNA damaging drugs

and radiotherapy

DNA damage response network

Supporting treatment optimisation

in the individual patient

Agent-based modelling

Agent based model

Simulation

A1 A2

A1Ai

A2

One-to-one mapping of cell components to computational agents

Agents at multiple levels:Protein, network motif, module (organelle, cell …)

Interaction rules

Translates wealth of molecular knowledge into component-based models

Patient-specific molecular data

?

TF1

S1 S2 SN

TF2 TFm

Signal-genetic network Environment

Transcription factors

Genes

DNA damage

Changes in genome activation

TF1

S1 S2 SN

TF2 TFm

Signal-genetic network Environment

Transcription factors

Genes

Agent-based modelling:

‘Agent’ (protein, motif, module) => behaviour rules

Kinetics/step function/Boolean variables

scale up to large networks

Challenge: Emergence

• Coherent behaviour of cells emerges from interactions between a large number of system components – proliferation, cell death, resistance to drugs

• ‘Computational’ definition of emergence: Unspecified properties and behaviours arise from interaction between agents rather than as a consequence of a single agent’s actions

• Methods for analysis needed e.g. for therapy target discovery

Detecting event patterns in time

• A simple event is a state transition due to a rule execution• A complex event is made up of a set of interrelated simple events

• Classification of complex events in a simulation allows one to discover associations between processes at different levels

• Published formalism available at www.cs.ucl.ac.uk/staff/C.Chen/research.html

Linking network simulations to integrated cell behaviour requires knowledge external to the simulation, the question of ‘biological meaning’

Challenge: ‘the gap’

A new synergy

• Data generation is still largely motivated by a non-systems-based research paradigm

• Systems biologists then seek to use these data to build and validate models of systems – with difficulties

• We need to rethink the relationship between experiment and modelling – both need to proceed within a complex systems framework – new kinds of experiments needed to investigate multi-level

relationships in the wet system• e. g., global signal network states need to be matched to cell-

level phenotypic measurements over time and under a range of conditions

• E-science systems modelling and experiment need to complement and synergise

Acknowledgements

• Nuno Rocha Nene (CoMPLEX PhD programme)• Chih-Chun Chen (interdisciplinary EPSRC DTA awards)• CR UK, Department of Health

• Published formalism available at www.cs.ucl.ac.uk/staff/C.Chen/research.html

• Decision support tool for ABM techniques www.abmsystemsbiology.info

• My email: s.nagl@ucl.ac.uk