Biological and Quantitative Models for Stem Cell Self-Renewal and Differentiation - Systems Biology Concepts and Insights - page 409

Chapter 22

Biological and Quantitative Models
for Stem Cell Self-Renewal and
Differentiation

Huilei Xu 1 , 3 , Dmitri Papatsenko 2 , 3 , Avi Ma'ayan 1 , 3 and Ihor Lemischka 2 , 3*
1 Department of Pharmacology and System Therapeutics, Mount Sinai School of Medicine, Systems Biology Center New York, One Gustave L. Levy
Place, New York, NY 10029, USA, 2 Department of Regenerative and Developmental Biology, Mount Sinai School of Medicine, One Gustave L. Levy
Place, New York, NY 10029, USA, 3 Black Family Stem Cell Institute, Mount Sinai School of Medicine, One Gustave L. Levy Place, New York, NY 10029,
USA

Chapter Outline
Current Problems and Paradigms in Stem Cell Research 427 Modular Network Design: from Kernels to Integrated

Models

Empirical vs. Systems Studies, Quantitative Biological
Models

433

427 Role of Transcription Regulatory Signals

and Transcriptional Gene Networks Deterministic and Stochastic View of Self-Renewal

and Differentiation

434

428 Dynamic Biological Reaction Model for the Core

Pluripotency Network Examples of Quantitative Models Explaining Stem

Cell Behavior

435

430 Model Validation and Overfitting 437

Models for Embryonic Stem Cells 430 Information Flow and Epigenetic Landscapes

in Differentiation Models for Hematopoietic Stem Cells 431 437

Strategies for Model Construction and Validation 431 Information Flow and Epigenetic Memory 437

Data Integration and Network Construction 431 Waddington Landscapes and Attractor States 438

Simple Binary Models for Complex Gene Networks 432 References 438

CURRENT PROBLEMS AND PARADIGMS
IN STEM CELL RESEARCH
Empirical vs. Systems Studies, Quantitative
Biological Models
Recent studies have revealed the potential of reprogram-
ming or transforming somatic cells into induced pluripotent
stem cells (iPS) [1

empirical studies, supported by bioinformatics and
genome-wide explorations [10,11] , but not much in part of
predictive biological models. Despite a constantly growing
reprogramming field, the low efficiency and the associated
expenses still impose barriers to therapeutic applications
[4,12,13] . Systems approaches, supported by quantitative
models, can bring new solutions and, possibly, solve the
current limits.
In general, the first step in model construction includes
broad integration of data, involving genome-wide expres-
sion or epigenetic studies [14,15] . Typically, these studies
identify the most prominent candidate genes selectively
expressed in self-renewing or differentiating stem cells.
Linking such candidate genes into networks is based on
their co-expression or the presence of similar binding
patterns for transcriptional regulators in the gene control

3] . However, the efficiency of reprog-
ramming remains low and the quality of the obtained cells
is often questionable [4] . The original retroviral reprog-
ramming method can transform only about 0.01% of
fibroblasts into iPS; more relevant to medicine, 'cleaner'
adenoviral reprogramming or direct delivery of reprog-
ramming proteins

e

into cells are even less efficient

(0.0001 0.001%) [5] . The original reprogramming 'cock-

tail' of four factors (Oct3/4, Sox2, c-Myc, and Klf4) and its
consequent modifications [6

e

9] were determined in rather

e