View metadata, citation and similar papers at core.ac.uk brought to you by CORE provided by Springer - Publisher Connector Open Access Software2006BoutrosetVolume al. 7, Issue 7, Article R66 Analysis of cell-based RNAi screens comment Michael Boutros*, Lígia P Brás†‡ and Wolfgang Huber† Addresses: *Signaling and Functional Genomics, German Cancer Research Center, Im Neuenheimer Feld 580, 69120 Heidelberg, Germany. †EMBL - European Bioinformatics Institute, Cambridge CB10 1SD, UK. ‡Centre for Chemical and Biological Engineering, IST, Technical University of Lisbon, Av. Rovisco Pais, P-1049-001 Lisbon, Portugal. Correspondence: Michael Boutros. Email: [email protected]. Wolfgang Huber. Email: [email protected] reviews Published: 25 July 2006 Received: 27 March 2006 Revised: 7 June 2006 Genome Biology 2006, 7:R66 (doi:10.1186/gb-2006-7-7-r66) Accepted: 25 July 2006 The electronic version of this article is the complete one and can be found online at http://genomebiology.com/2006/7/7/R66 © 2006 Boutros et al.; licensee BioMed Central Ltd. This is an open access article distributed under the terms of the Creative Commons Attribution License (http://creativecommons.org/licenses/by/2.0), which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited. reports Analysis<p>cellHTS of cell-based is a new method RNAi screens for the analysis and documentation of RNAi screens.</p> Abstract RNA interference (RNAi) screening is a powerful technology for functional characterization of deposited research biological pathways. Interpretation of RNAi screens requires computational and statistical analysis techniques. We describe a method that integrates all steps to generate a scored phenotype list from raw data. It is implemented in an open-source Bioconductor/R package, cellHTS (http:// www.dkfz.de/signaling/cellHTS). The method is useful for the analysis and documentation of individual RNAi screens. Moreover, it is a prerequisite for the integration of multiple experiments. refereed research refereed Rationale However, the identification of genes whose mutation leads to RNA interference (RNAi) is a conserved biological mecha- an altered phenotype can be cumbersome and slow. Rapid nism to silence gene expression on the level of individual reverse genetics by RNAi allows the systematic screening of a transcripts. RNAi was discovered in Caenorhabditis elegans whole genome whereby every single transcript is depleted by when Fire and Mello [1] observed that injecting long double- siRNAs or dsRNAs. Genes with unknown functions can then stranded (ds) RNAs into worms led to efficient silencing of be classified according to their phenotype. The speed of homologous endogenous RNAs. Subsequent studies showed reverse genetic screens using high-throughput technologies that the RNAi pathway is conserved in Drosophila and verte- promises to accelerate significantly the functional characteri- brates, and can be used as a tool to downregulate the expres- zation of genes [7]. RNAi screens have been successfully used interactions sion of genes in a sequence specific manner [2,3]. Long in C. elegans to elucidate whole organism phenotypes and for dsRNAs are commonly used in Drosophila and C. elegans. In cell-based assays in fly, mouse, and human cells [8-17]. Fig- mammalian cells, long dsRNAs induce an interferon ure 1 outlines the main steps in cell-based high-throughput response, and therefore short 21 mer RNA duplexes (small screening (HTS) experiments. interfering RNAs [siRNAs]) are effective in silencing target mRNAs [4,5]. The analysis of data sets generated by high-throughput phe- notypic screens poses new methodological challenges. The Cell-based RNAi screens open new avenues for the systematic richness of phenotypic results can range from single numeri- information analysis of genomes. Traditionally, genetic screens by ran- cal values to multidimensional images from automated dom mutagenesis have been successful in identifying and microscopy. Whereas analysis of functional genomic datasets characterizing genes in model organisms that are required for generated by transcriptome and proteome analysis has specific biological processes [6]. These led to the discovery of attracted considerable interest, analysis of high-throughput many pathways that were later implicated in human disease. cell-based assays has lagged behind. Each study has been con- Genome Biology 2006, 7:R66 R66.2 Genome Biology 2006, Volume 7, Issue 7, Article R66 Boutros et al. http://genomebiology.com/2006/7/7/R66 Genome annotation RNAi Library annotation file library design Screen description file Cell-based assay format Plate list file Plate configuration file Large-scale experiment Screen data files Screen logfiles Computational analysis Compendia and web reports ExperimentalFigure 1 steps in a cell-based HTS assay Experimental steps in a cell-based HTS assay. A cell-based HTS assay consists of a set of experimental steps, shown in the left part of the figure, which are recorded in a set of corresponding data structures, shown in the right part of the figure. HTS, high-throughput screening. Genome Biology 2006, 7:R66 http://genomebiology.com/2006/7/7/R66 Genome Biology 2006, Volume 7, Issue 7, Article R66 Boutros et al. R66.3 ducted using unique custom-tailored analytical methods. Example data Although this may be appropriate within the context of a sin- We demonstrate the analysis methodology using a published gle study, it makes the integration or comparison of datasets example dataset from a genome-wide RNAi screen for dsR- comment difficult if not impossible. The documentation and minimal NAs that cause cell viability defects in cultured Drosophila information required for reporting RNAi experiments remain cells [9]. In these experiments, Kc167 cells were treated with unresolved issues [18]. Nevertheless, as the number of RNAi dsRNAs from a library consisting of more than 20,000 dsR- screens performed by different groups increases, it will be NAs. After 5 days cell viability was determined using a lumi- instrumental that reliable tools are developed for their inte- nescence readout by a microplate reader. The library was gration and comparative analysis. provided in an arrayed format, in which each location in a 96- well or 384-well microplate uniquely identifies the dsRNA. We present a software package for the construction of analy- The cell viability screen was performed in duplicate, and raw reviews sis pipelines for genome-wide RNAi screens. Step by step, it results are available as plate reader outputs containing rela- leads from raw data files to annotated phenotype lists and tive luminescence readings. Details of the screening proce- documentation (Figure 2). Comprehensive data visualization dure are described elsewhere [9], sequence information is and quality control plots aid in identifying experimental out- available from our website [19], and the data are provided as liers. The data can be normalized for systematic technical var- part of the examples in the documentation of the cellHTS iations, and statistical summaries are calculated. Quality package. The analysis we present here generally follows the metrics of the experiment help in assessing the strength of the analysis performed for the original report [9]. results. The complete analysis is documented as a computer- reports readable living document. A navigable presentation of the Additionally, we provide a sample dataset of a dual channel results is produced as a set of HTML pages that is amenable, experiment. This type of experimental design is used to meas- for example, for provision as supplemental information ure, for instance, the phenotype of a pathway-specific alongside publication of the study. reporter gene against a constitutive reporter that can be used for normalization purposes. Typical examples for such exper- imental setups are dual-luciferase assays, whereby both fire- fly and Renilla luciferase are measured in the same well. In deposited research principle, multiplex assays can consist of many more than two channels, such as in the case of flow-cytometry readout [20] or other microscopy-based high-content approaches. Import raw data files Data import and assembly In this section we discuss the information that is necessary to Per plate quality control describe a cell-based HTS experiment. In addition to the pri- research refereed mary data files, descriptions of the experimental setup, the configuration of screening plates, and annotations for the RNAs need to be provided. A schematic representation of a Data normalization Documentation screening setup and the corresponding files is shown in Fig- of RNAi screening ure 1. The input data consist of several tabular files: the anno- and data processing tation of the library, a screen description file, a plate list file, steps a plate configuration file, the primary data, and - if available Scoring of phenotypes - a log file of the screening procedure. interactions The screen description file contains a general description of Annotation and analysis the screen, its goal, the conditions under which it was per- formed, references, and any other information that is impor- tant for the analysis and biological interpretation of the experiment. The purpose of this file is similar to that of the Export as HTML report and compendia experiment design section of a MIAME-compliant dataset [18]. information The plate configuration file contains information about the AnalysisFigure 2steps for a cell-based