MIQE précis: Practical implementation of minimum standard guidelines for fluorescence-based quantitative real-time PCR experiments
© Bustin et al; licensee BioMed Central Ltd. 2010
Received: 13 September 2010
Accepted: 21 September 2010
Published: 21 September 2010
The conclusions of thousands of peer-reviewed publications rely on data obtained using fluorescence-based quantitative real-time PCR technology. However, the inadequate reporting of experimental detail, combined with the frequent use of flawed protocols is leading to the publication of papers that may not be technically appropriate. We take the view that this problem requires the delineation of a more transparent and comprehensive reporting policy from scientific journals. This editorial aims to provide practical guidance for the incorporation of absolute minimum standards encompassing the key assay parameters for accurate design, documentation and reporting of qPCR experiments (MIQE précis) and guidance on the publication of pure 'reference gene' articles.
Fluorescence-based quantitative real-time PCR (qPCR) is commonly regarded as a straightforward, mature technology and has become a ubiquitous mainstay of molecular biology. However, obtaining, analysing and interpreting qPCR data is not a trivial issue. The progressive relegation of qPCR materials and methods to online supplements is resulting in an increasing tendency for scientific publications to contain insufficient technical information for that work to be reproduced; frequently that information is lacking altogether. Equally disconcertingly, the modest details included with publications frequently reveal the use of flawed practices that are likely to result in the publication of erroneous conclusions. This often makes it impossible for reviewers and readers to judge a manuscript's technical adequacy, whilst impeding journal editors' ability to reach publication decisions based on its validity and coherence.
As sample handling affects experimental variability, it is essential to report tissue source and to provide information about conditions of storage, particularly if samples have been obtained during the course of a longitudinal study stretching over several years. Any deviations from manufacturers' protocols must be recorded. For RNA templates, the extent of residual genomic DNA contamination must be reported, ideally for each sample, or if numerous samples are being used, for a representative set of samples, by comparing quantification cycles (Cq) obtained with and without reverse transcription (RT) for each nucleic acid target. Where RNA quantities are ample for accurate RNA measurement, both quantity and quality of extracts should be recorded. Although no perfect quality assessment methods exist, measurements such as those obtained from microfluidics-based systems or 3':5'-type assays can serve as basic indicators of RNA integrity . Authors should not quantitatively compare RNAs of widely dissimilar quality, e.g., RIN/RQI values of 4.5 versus 9.5, and be aware that rRNAs yielding similar RIN/RQI numbers can contain mRNAs that differ significantly in their integrity. Every sample or, in the case of very large numbers, representative samples, should also be tested for the absence of inhibitors using either an "alien" spike or a dilution series of target genes.
Database accession numbers, amplicon size and primer sequences must be reported for each target. We would recommend that probe sequences and the identities, positions, and linkages of any dyes and/or modified bases are also reported. We encourage the use of validated assays, e.g. those available from RTPrimerDB, as this helps with assay standardisation and does not require extensive validation. For newly reported assays, the use of primer design software is encouraged. Primer specificity should be validated in silico (BLAST specificity analysis) and empirically (ideally by DNA sequencing, but certainly using gel electrophoresis or melting profiles). Evidence for the optimisation of primer and MgCl2 concentration, as well as that of the annealing temperature should be provided and the PCR efficiency of each assay should be reported. This is most conveniently performed by means of calibration curves, which must be included for review, although only slopes, y intercepts, linear dynamic range and LOD need be included with the publication. A measure of intra-assay variation (repeatability) and for diagnostic assays inter-assay precision (reproducibility) between sites and different operators should also be reported. For multiplexing it is also essential to demonstrate that accurate quantification of multiple targets in a single tube is not impaired.
Reverse transcription qPCR (RT-qPCR)
no template controls (NTC). These are essential for detecting PCR contamination and must be performed with every experiment. They help establish conditions for data rejection from assays utilising SYBR Green chemistry, where primer dimers may result in positive Cqs being recorded in the NTCs. NTCs should be the last samples to be dispensed and loaded.
positive controls. These are most conveniently performed as target gene-specific dilution curves and are essential if PCR measurements are of low copy numbers and where it is essential to have confidence in a negative result, for example when detecting pathogens.
no amplification controls (NAC). These provide valuable information when using probes, as they help monitor any probe degradation.
We encourage the use of specialist software for data analysis, confidence estimation and outlier identification. The handling of such data must be specified as well as the statistical methods used to evaluate assay precision. Information relating to experimental lay-out must be supplied: we encourage the use of a sample maximisation strategy, i.e., running as many samples as possible in the same run, as opposed to a gene maximisation strategy that analyses multiple genes in the same run. This is because it minimises any technical, run-to-run variation between different samples for the comparison of gene expression levels. If not all samples can be analysed in the same run, identical samples that are tested in both runs ("inter-run calibrators" or IRCs) must be analysed. Measuring the difference in Cq or the normalised relative quantity between the IRCs in different runs allows the calculation of a correction or calibration factor to remove run-to-run differences .
Normalisation with reference genes
mRNA data are most commonly normalised using reference genes as internal controls. Their utility must be validated experimentally for particular tissues or cell types using the specific experiment in question. Unless fully validated single reference genes are used, normalisation should be performed against multiple reference genes, chosen from a sufficient number of candidate reference genes tested from independent pathways using at least one algorithm (e.g., GeNorm). In general, the use of fewer than three reference genes is not advisable and reasons for choosing fewer must be specifically addressed.
For large-scale miRNA expression profiling experiments normalisation should be performed against the mean expression value of all expressed miRNAs. For small scale experiments, pilot experiments analogous to those used for mRNA reference gene selection should be used to identify suitable reference miRNAs. It is worth noting that miRNAs isolation should be carried out using specialised extraction protocols, as the efficiency of miRNA extraction is reagent-dependent and can be variable.
Demonstrate the use of the selected reference genes for further investigations, e.g., to normalize RT-qPCR data of target genes as part of a larger study
Introduce and validate a novel approach for reference gene identification or normalisation, e.g., algorithms or software
Highlight rare cases or a standardized experimental setup of outstanding importance.
However, 'reference gene papers' that do not meet the criteria outlined here may still be suitable for peer review in other journals such as http://www.biomedcentral.com/bmcresnotes/. These papers are available and will continue to be collated in a topical series (http://www.biomedcentral.com/bmcresnotes/topicalseries/qRT-PCR_normalization_optimization).
Checklist for authors' of MIQE précis at time of manuscript submission, detailing information about individual parameters associated with each step of the RT-qPCR workflow.
If cancer, was biopsy screened for adjacent normal tissue?
Method of preservation
Storage time (if appropriate)
If using samples >6 months old
Intron-spanning primers/no RT control
Method of testing
exon location, amplicon size
even if previously published
identify LNA or other substitutions
primer concentration/annealing temperature
Linear dynamic range
spanning unknown targets
Limits of detection
LOD detection/accurate quantification
copy numbers not Cq
detailed description, concentrations, volumes
supplier, Lot number
Cq & melt curves
ΔCq beginning:end of qPCR
e.g., biological replicates
Transparent, validated normalisation
e.g., GeNorm summary
It is becoming increasingly obvious that it is no longer possible to assess the technical adequacy of qPCR-based peer-reviewed publications by scrutinising either their materials and methods or online supplements sections. We propose the above guidance to uphold the quality of research published in BMC Molecular Biology and to help standardise the way in which qPCR-based experiments and data are reported.
- Bustin SA, Benes V, Garson JA, Hellemans J, Huggett J, Kubista M, Mueller R, Nolan T, Pfaffl MW, Shipley GL, et al: The MIQE guidelines: minimum information for publication of quantitative real-time PCR experiments. Clin Chem. 2009, 55 (4): 611-622. 10.1373/clinchem.2008.112797View ArticlePubMedGoogle Scholar
- Nolan T, Hands RE, Bustin SA: Quantification of mRNA using real-time RT-PCR. Nature Protocols. 2006, 1 (3): 1559-1582. 10.1038/nprot.2006.236View ArticlePubMedGoogle Scholar
- Derveaux S, Vandesompele J, Hellemans J: How to do successful gene expression analysis using real-time PCR. Methods. 2010, 50 (4): 227-230. 10.1016/j.ymeth.2009.11.001View ArticlePubMedGoogle Scholar
This article is published under license to 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.