Inference of Boolean Networks from Time Series Data with Realistic Characteristics

Author

Erkkilä, Timo ; Korpelainen, Tomi ; Yli-Harja, Olli

Author_Institution

Tampere Univ. of Technol., Tampere

fYear

2007

fDate

10-12 June 2007

Firstpage

1

Lastpage

4

Abstract

In this paper, we infer Boolean networks based on simulated data. Depending on the approach, the simulated data is either extracted from Boolean networks or from a biochemical network model. Noise was added using the existing hierarchical error model (HEM). We use Boolean inference algorithms based on best-fit to find the connections between nodes. In order to validate the inference processes, we compare the ground truth and inferred connections. The results show that the best-fit inference algorithm is suitable for Boolean approach, but it fails when using the biochemical network model.

Keywords

biochemistry; biology computing; inference mechanisms; noise; physiological models; time series; Boolean inference algorithm; Boolean networks; best-fit inference algorithm; biochemical network model; hierarchical error model; noise; time series; Biological system modeling; Boolean functions; Data analysis; Data mining; Filtering; Inference algorithms; Interpolation; Mathematical model; Signal processing algorithms; State-space methods;

fLanguage

English

Publisher

ieee

Conference_Titel

Genomic Signal Processing and Statistics, 2007. GENSIPS 2007. IEEE International Workshop on

Conference_Location

Tuusula

Print_ISBN

978-1-4244-0998-3

Electronic_ISBN

978-1-4244-0999-0

Type

conf

DOI

10.1109/GENSIPS.2007.4365848

Filename

4365848