Recent advances and developments in genomics and functional genomics have enabled the development of high-throughput experiments and assays that are systematically applied to screen for effects induced by systematic perturbations of cells via the underexpression or overexpression of the respective RNA/proteins in living cells. However, RNA/Protein over- and under-expression are by far not the only means to perturb cells but perturbations can be administered in several ways: Small chemical compounds, physicochemical parameters (temperature, humidity, supplements to the medium), and other treatments are utilized.

The amount of data and information from cellular assays (e.g. RNAi) rapidly growing. The evaluation, comparison and integration of results from different cellular assays being carried out in different laboratories necessitates a comprehensive description of the experimental conditions that is provided in a commonly accepted format, as well as standardized output formats for variable types of experimental data.
The rationale, materials, growth conditions prior to, during, and after the perturbation, as well as all experimental processes need to be recorded and documented in order to fully describe the set-up and progress of a cellular assay project, and to be able to understand and follow the data analysis and knowledge generation processes. Standardized nomenclatures/ontologies should be employed wherever possible.

However, quality management and stardardization (QM&S) for high-throughput cellular assays is still in its infancy.
We have initiated the establishment of such QM&S measures as a close collaboration with other NGFN partners. We have implemented bioinformatics tools and databases that are suitable to analyze, store and integrate the different types of data from high-throughput cellular assays and to include further annotation that is based on external information. We use a Laboratory Information Management System (LIMS) that allows process tracking during the assay.

We are continually optimizing the work flows and protocols in the field of cellular functional assays. 

One illustrating example:
Quantitative RT-PCR is a most relevant means of quality control within the functional assays of SMP-Cell, and used for instance to validate RNAi-induced gene knock-down. However, this fast and well-established method is still expensive when performed at a large scale. We are currently evaluating the reproducibility of the Exiqon qRT-PCR probe set, which covers most known genes with only 90 probes, and is a potentially cost-effective alternative to standard Taqman analysis.
Furthermore, we tested enzymatically-generated siRNAs and chemically-generated siRNAs from three different suppliers for their functionality in three different cell lines. Quality of knock-down was measured via qRT-PCR of the target genes/mRNAs. Cell-type and concentration dependent effects and preferences were observed with the different siRNA resources. RNAi appears to be an extremely valuable technology, however there seem to be some challenges that require further work to be carried out in order to understand the mechanisms and principles.

In the frame of an international initiative, we propose the definition of minimum information guidelines – MIACA (Minimum Information About a Cellular Assay)– [MS-Word][PDF] for the standardized description of cell based assay projects. A number of closely related working parties covering both large scale cell-based assays and bioinformatics for data analysis are currently working on the definition of the minimal requirements to describe a cell based assay project.

MIACA will be compatible with the already existing formats developed by the HUPO Proteomics Standards Initiative and utilizes established ontologies, which will enable the user to describe related information in a consistent way. Structures and ontologies should be used as widely as possible from existing standards (e.g. OBO - http://obo.sf.net/, MGED - http://mged.sf.net/, HUPO-PSI - http://psidev.sf.net/, FuGE - http://fuge.sf.net/, more universal features should be taken from general standards (e.g. SI-units).

Information is structured in a modular way that enables useful querying, automated data analysis and efficient data mining. Current modules are 

Study design
Administrative details (institute & scientist)
Experimental motivation (project description)
Description of perturbator type(s)
Origin of cell line used – ATCC, other, primary cells,…
Materials & reagents (array type, media)
Instruments (data acquisition)

2.  Pre-treatment of cells prior to the perturbation(s)
Conditions of cell culture, passage number, number of cells, …
Other actions as washes, trypsinization, …

3- Perturbation
Administration of perturbator(s), which may be siRNAs, expression constructs, small molecules, a shift of temperature, …
Conditions under which the cells are subjected to the perturbation(s) (e.g. time)

4. Post-treatment of cells
Conditions following the perturbation. Conditions of cell culture (chase experiment) prior to fixation and data acquisition
Post-treatment to prepare samples for data acquisition (fixation, antibody incubation, washes, …)

5. Data acquisition
The process of data acquisition (Plate reader, FACS, High content screening microscopy)
Instrument settings (e.g. excitation and emission wavelengths with filter sets, lamp energy, time for data acquisition)

6. Data analysis
Procedures applied for the filtering and normalization of data (based on which controls).
The Scoring schema and statistical analysis procedures for the production of a “hit-list”.

Information should be of sufficient detail to allow for the interpretation of the validity of the project and its outcome, to allow for comparisons with similar projects, and (in principle) to permit replication.  When necessary, additional ontologies need to be developed/adopted. Otherwise, 'open' controlled vocabularies could be used (for referencing external ontologies or the inclusion of values as name/value/source triplets (as described e.g. for MIAME and PSI MIF).

Respective documentation (initial draft of the standard, an annotated XML Schema and sample files) can be found here.

Website of the SMP Cell involved in this NGFN quality management project: http://www.dkfz.de/smp-cell/cell.org/

Homepage