iConcept Press Logo
Email Password Remember me
Genomics II: Bacteria, Viruses and Metabolic Pathways
Title
Genomics II: Bacteria, Viruses and Metabolic Pathways
Editor
iConcept Press
Price
USD$139.00
ISBN
978-1-480254-145
Clicks
22866

Chapter 1

Genomics II: Bacteria, Viruses and Metabolic Pathways

Statistical Analysis and Processing of Cellular Assays

by Bettina Knapp and Lars Kaderali

Viewed: 2541

Abstract

Recent technical developments in the fi eld of cellular assays have allowed for qualitative and quantitative improvements of data generation. This results in facilitated measurements which enable that biological experiments such as RNA interference (RNAi) are performed in a high-throughput, high-content manner. Thereby, a better understanding about complex biological systems such as viral life cycles can be gained. Yet, even when studying the same virus, already published screens show only a very small overlap [1]. This suggests that the devil may be in the details and that the noisy data has to be analyzed using sophisticated methods. To extract the true biological signal, the data has to be unraveled from the various artifacts such as pipetting, robot or scanning failures as well as from the diff erences due to screens performed with di fferent experimental conditions in diff erent labs. Snijder and colleagues showed that the population context greatly influences each cell\\\'s phenotype on cellular assays [2]. However, most studies are not using single-cell measurements for cellular assays but summarize the phenotype of the cells treated in the same way. In addition, only some standard methods are used for the data analysis and statistical hit scoring. Therefore, we introduce a new data analysis approach for cellular assays which is based on single-cell measurements. Each individual cell is characterized by several cell-context features such as cell size, shape and local cell density. Assuming that each cell\\\'s context is influencing the phenotype, we use the individual features to normalize the data. Additionally, technical features are computed for each cell to account for the technical artifacts given in the data. This results in a multivariate feature space of several thousands or millions of cells. For the normalization of the single-cell measurements against these features we use multivariate adaptive regression splines [3] which is a non-parametric regression technique producing continuous models in a greedy heuristic approach. The hit scoring on these normalized individual cell measurements is done using a method which is based on ideas from gene set enrichment analysis. In total, the single-cell analysis allows to exploit the high-content data in more detail and to account for phenotypic eff ects due to noise and experimental differences. Thereby, the reliability and reproducibility of the results is improved and false positive and false negative hits are minimized. References [1] S. Cherry, \\\"What have RNAi screens taught us about viral-host interactions?,\\\" Curr. Opin. Microbiol., vol. 12, pp. 446{452, Aug 2009. [2] B. Snijder, R. Sacher, P. Ramo, E. M. Damm, P. Liberali, and L. Pelkmans, \\\"Population context determines cell-to-cell variability in endocytosis and virus infection,\\\" Nature, vol. 461, pp. 520{523, Sep 2009. [3] J. H. Friedman, \\\"Multivariate adaptive regression splines,\\\" The annals of statistics, vol. 19, no. 1, pp. 1-141, 1991.

Author Details

Bettina Knapp
Institute for Medical Informatics and Biometry, Medical Faculty Carl Gustav Carus, Technische Universität Dresden, Germany, Germany
Lars Kaderali
Institute for Medical Informatics and Biometry, Medical Faculty Carl Gustav Carus, Technische Universität Dresden, Germany, Germany

Citation

Bettina Knapp and Lars Kaderali. Statistical Analysis and Processing of Cellular Assays. In Genomics II: Bacteria, Viruses and Metabolic Pathways. ISBN:978-1-480254-145. iConcept Press. 0000.

Download