Encyclopedia of Systems Biology

2013 Edition
| Editors: Werner Dubitzky, Olaf Wolkenhauer, Kwang-Hyun Cho, Hiroki Yokota

Prediction Model Integration

Reference work entry
DOI: https://doi.org/10.1007/978-1-4419-9863-7_238



In the biomedical area an integrative predictive model is one that combines different biomarkers obtained by different means and apply them with a statistical model to select specific treatments for individual patients based on patient’s characteristics, disease status, and clinical outcome (George 2008).

In general terms, a predictive model aims to capture statistical interactions between biomarkers and treatments. It is commonly abstracted and designed from empirical data (i.e., clinical and omic data). Predictive models should be statistically validated to assess their goodness-of-fit to empirical data, and clinically validated to verify their applicability.

The complexity of biological processes frequently necessitates the use of complex statistical models with a large number of variables. Against this complexity, the amount of biomedical data available is often not sufficient to fully determine a statistical...

This is a preview of subscription content, log in to check access.


  1. George SL (2008) Statistical issues in translational cancer research. Clin Cancer Res 14(19):5954–5958PubMedGoogle Scholar

Copyright information

© Springer Science+Business Media, LLC 2013

Authors and Affiliations

  1. 1.Institute of Biological, Environmental and Rural SciencesAberystwyth UniversityCeredigionUK
  2. 2.School of Computing SciencesUniversity of East AngliaNorwichUK