Article 5: qPCR data analysis - amplification plots, Cq and normalization (2023)

Posted: October 9, 2009 | Tania Nolan, Global Director of Applications and Technical Support, Sigma-Aldrich. Stephen Bustin, Professor of Molecular Science, Center for Academic Surgery, Institute of Cellular and Molecular Science, Barts and London School of Medicine and Dentistry and Jim Huggett, Scientific Director of Molecular and Cellular Biology, LGC |no comments yet

A key advantage of qPCR technology is its apparent lack of complexity. an assay consisting of a simple procedure of combining oligonucleotides, PCR mastermix buffer and template nucleic acid to produce a qPCR reaction is considered non-demanding. This practical simplicity is complemented by the absence of any post-analysis handling requirement, as well as the development of user-friendly data analysis software that makes the generation and visualization of data in the form of amplification plots extremely simple. However, as we mentioned in the first four articles in this series, translating an attractive graphical enhancement into accurate and meaningful data is far from trivial and requires a number of additional considerations.

A key advantage of qPCR technology is its apparent lack of complexity. an assay consisting of a simple procedure of combining oligonucleotides, PCR mastermix buffer and template nucleic acid to produce a qPCR reaction is considered non-demanding. This practical simplicity is complemented by the absence of any post-analysis handling requirement, as well as the development of user-friendly data analysis software that makes the generation and visualization of data in the form of amplification plots extremely simple. However, as we mentioned in the first four articles in this series, translating an attractive graphical enhancement into accurate and meaningful data is far from trivial and requires a number of additional considerations.

A key advantage of qPCR technology is its apparent lack of complexity. an assay consisting of a simple procedure of combining oligonucleotides, PCR mastermix buffer and template nucleic acid to produce a qPCR reaction is considered non-demanding. This practical simplicity is complemented by the absence of any post-analysis handling requirement, as well as the development of user-friendly data analysis software that makes the generation and visualization of data in the form of amplification plots extremely simple. However, as we mentioned in the first four articles in this series, translating an attractive graphical enhancement into accurate and meaningful data is far from trivial and requires a number of additional considerations.

Incorporating the recommendations from the previous articles into standard operating procedure will go some way to creating amplification plots that are not only aesthetically pleasing, but are the results of efficient qPCR assays and accurately relate to the original nucleic acid sample. However, the process of analyzing this data is susceptible to misinterpretation and should be approached with the same caution as during wet lab procedures. Although the automatic settings included in the instrument software are useful and can prevent some wild errors, it is recommended that you use common sense and discretion in performing data analysis procedures. Adopting a logical, systematic and consistent approach to data analysis will reduce the scope for misinterpretation, thereby reducing the scope for erroneous conclusions and misreporting.

Examination of the reinforcement diagram

The physical appearance of the gain plot provides an initial indication of the quality of the analysis. Initially, some experience is required to become familiar with the units of the fluorescence scale used by the instrument. Consider the choice of values ​​for the scale of the Y-axis, which indicates fluorescence: the instruments are different and the absolute values ​​will differ from each other (see Figure 1). The initial focus for data analysis should be on raw fluorescence values, baseline settings, thresholding (if used) and melting curves (eg for SYBR Green I, Molecular Beacons or Scorpions). Analysis of the raw data reveals whether the fluorescence intensity range is appropriate. The raw signal in the exponential gain region used to determine the quantization cycle (Cq)1 must be well above background and well below saturation to avoid problems with peaking and poor signal uniformity. However, raw fluorescence data must be further analyzed to eliminate inter-instrument variability and ensure that Cq values ​​can be directly compared (see Figure 2). One approach is to subtract the background signal from all traces so that all baseline data are set to zero (dR). To do this, the software needs to identify the data that makes up the background noise before actually amplifying it. Figure 3 shows the data before baseline normalization and Figure 4 shows the data after baseline normalization. The significant increase in fluorescence during exponential amplification (blue trace) results in amplification plots that can be analyzed, while the low signal data from the red traces results in normalized data that is sharp due to the low signal-to-noise ratio. Several baseline normalization algorithms are available, the most powerful being those that examine each gain plot individually and set a baseline for each plot accordingly. In addition, it is useful to have the ability for the user to change these settings. In instruments that have non-uniform detection systems, the data require further correction to compensate for differences in fluorescence intensity across the block. This correction is traditionally achieved by including a fixed reference dye concentration (eg, ROX) and plotting the relative fluorescence intensity of the signal of interest for the fixed reference dye against the number of cycles (dRn/ΔRn/RFU).

Setting a boundary

An interesting exercise (worth moments of consideration on an otherwise boring Friday afternoon) is to construct a standard curve spanning a dilution range of six logarithms, preferably with only one to five target replicates at the highest dilution, then run the qPCR and read the Cq values ​​recorded using the instrument's automatic settings . Now set the threshold higher and then lower and get the Cq readings for each threshold setting. Threshold setting naturally affects the value of Cq. More important than the absolute values ​​of Cq are the differences between the values ​​for different concentrations. Calculate the differences between the following dilution points for different threshold settings. Are they identical? If the analysis is well optimized and linear, you will notice that the absolute values ​​of Cq vary with the threshold setting, but the relative Cq (ΔCq) remains almost identical. This is because the amplification plots for each dilution of the sample are parallel and therefore a change in the threshold affects each amplification plot to the same extent. However, determinations are not always so reliable, and examples of plots of higher dilution factors (higher Cq) that have a different slope than amplification plots of lower dilutions (lower Cq) are widely found. In these cases, the absolute threshold setting affects the estimation of the amount or relative amount of data. The threshold should be set at the logarithmic phase of the gain and at the point where all gain plots are parallel, bearing in mind that quantification of those that are not parallel leads to an unacceptably large error. It is important to use a common-sense approach to examining the effect of amplification and problems arising from poorly designed dilution curves. This is shown in a publication that criticizes a peer-reviewed paper for reporting incredibly high amplification rates and applying incorrect statistical analyzes that call into question the reliability and relevance of conclusions based on these findings2. Recently, instrument manufacturers such as BioRad have introduced alternative methods for data analysis. These include innovations such as those that apply a multivariate, nonlinear regression model to individual well traces and then use that model to calculate an optimal Cq value, trying to avoid some of the problems mentioned above. There are many algorithms in development and only time will tell if they provide a reliable solution. Without some reference concept, Cq values ​​are limited values ​​due to the fact that they can be altered by buffer, primers, machines, fluorescence chemistry along with many other factors.

Export Cq or quantity values

As described above, the value of Cq alone is of limited use since the absolute value depends entirely on the threshold setting in the same experiment and on a number of other factors when experiments are compared. Therefore, it usually does not make sense to publish a Cq3 value. Instead, for most studies, the Cq value must be converted to a more informative unit of measurement. In some cases, an estimate of relative target abundance can be determined by examining the difference in Cq values ​​between samples analyzed by the same qPCR assay. As shown above, this must be done with caution since the use of Cq differences assumes that increasing all target concentrations results in gain plots that are parallel to each other and shifted along the X-axis in a manner proportional to the initial concentration. Alternatively, serial dilution of a target concentration of known or relative concentration can be used to calibrate target concentration measurements in test samples3,4.

Using either of these approaches alone to quantify a target in a sample could result in inaccurate data because they rely on a number of assumptions. These include assumptions of similar sample quality, that any degradation does not affect quantification, equal reverse transcription (RT) efficiency, absence or equal effects of reaction inhibitors or enhancers5, equal loading of gDNA or cDNA in the qPCR assay, and equal PCR efficiency in each sample for each target. Since it is clear that none of these assumptions are necessarily valid, there must be a procedure to compensate for these variables.

Normalization

Η κανονικοποίηση δεδομένων στοχεύει στην αντιμετώπιση ορισμένων από τις ελλείψεις που αναφέρονται παραπάνω, και υπάρχουν πολλές διαφορετικές προσεγγίσεις, με νέες προτεινόμενες συνεχώς. Η κανονικοποίηση στη συνολική μάζα ή όγκο δείγματος είναι μια κληρονομική προσέγγιση που απομένει από τις τεχνικές Northern blotting όταν η φόρτωση γέλης και η μεταφορά δείγματος σε διηθητικό χαρτί επικυρώθηκε με ανίχνευση για rRNA ή τα λεγόμενα γονίδια housekeeping όπως το GAPDH, των οποίων η έκφραση υποτίθεται ότι είναι σταθερή μεταξύ ατόμων , πειραματικές συνθήκες ή φυσιολογικές καταστάσεις. Ενώ εξακολουθεί να χρησιμοποιείται για μετρήσεις RT-qPCR, υπάρχουν αρκετά μειονεκτήματα στη μέτρηση των επιπέδων mRNA ή miRNA σε σχέση με τη μάζα ή τον όγκο του δείγματος. Για παράδειγμα, σε σύγκριση δειγμάτων διαφορετικής προέλευσης π.χ. βιοψίες όγκου και φυσιολογικού ιστού είναι εσφαλμένο να υποθέσουμε ότι η ίδια μάζα ιστού περιέχει παρόμοιους αριθμούς κυττάρων ή ότι η σχετική κατανομή των πολλαπλασιαζόμενων κυττάρων είναι ίση. Η κανονικοποίηση έναντι του ολικού RNA θα υποτιμήσει την έκφραση των γονιδίων-στόχων στις βιοψίες όγκου. Αυτή η προσέγγιση μπορεί να είναι πιο κατάλληλη όταν τα δείγματα εξάγονται με χρήση μικροτομής σύλληψης με λέιζερ και στοχεύονται ένας ακριβής αριθμός και παρόμοια κύτταρα. ακόμη και τότε αυτή η προσέγγιση δεν είναι ιδανική. Μια σχετική τεχνική, που θυμίζει και πάλι το Northern blotting, είναι η κανονικοποίηση σε συγκέντρωση DNA ή RNA. Ενώ η μέτρηση του αριθμού αντιγράφων γονιδίου σε σχέση με τη συγκέντρωση του εισαγόμενου DNA είναι μια απόλυτα έγκυρη προσέγγιση, η κατάσταση είναι πιο περίπλοκη για την ποσοτικοποίηση μεταγραφής. Όταν προσδιορίζεται η συγκέντρωση RNA, η συντριπτική πλειοψηφία του συστατικού RNA είναι ριβοσωμικό RNA (rRNA). Η μεταγραφή του rRNA είναι ανεξάρτητη από τη μεταγραφή του αγγελιαφόρου RNA (mRNA) αφού μεταγράφεται από διαφορετικά ένζυμα. Επιπλέον, καθώς το rRNA αποτελεί το ~80% του κλάσματος RNA, η κανονικοποίηση σε rRNA θα κάλυπτε ανεπαίσθητες αλλαγές στο συστατικό mRNA. που τυπικά περιλαμβάνει 2-5%. Επιπλέον, αυτή η προσέγγιση δεν λαμβάνει υπόψη παραλλαγές στην ποιότητα RNA του εκμαγείου ή αλλαγές στα επίπεδα rRNA που εξαρτώνται από την κατάσταση κυτταρικής διαφοροποίησης/πολλαπλασιασμού. Ωστόσο, αν και δεν είναι ιδανικές, μπορεί να υπάρχουν καταστάσεις όπου δεν υπάρχουν άλλες εναλλακτικές από τη μέτρηση σε σχέση με το συνολικό RNA και ορισμένα πακέτα ανάλυσης όπως το λογισμικό GenEx από το MultiD6 επιτρέπουν την αξιολόγηση της συνολικής συγκέντρωσης RNA ως τεχνική κανονικοποίησης μαζί με άλλες προσεγγίσεις. Μια θεωρητική λύση θα ήταν ο καθαρισμός του mRNA και η κανονικοποίηση έναντι του ολικού mRNA. Δυστυχώς, η διαδικασία καθαρισμού εισάγει ανακρίβειες και ένα επιπλέον στάδιο επεξεργασίας που είναι ανεπιθύμητο και σε πολλές περιπτώσεις η βιοψία είναι πολύ μικρή για να επιτρέψει τον αποτελεσματικό καθαρισμό του re mRNA. Μια πολύ κοινή προσέγγιση για τη διόρθωση των διαφορών στα δείγματα είναι η έκφραση της συγκέντρωσης στόχου σε σχέση με αυτή ενός μόνο εσωτερικού γονιδίου αναφοράς. Για να είναι έγκυρη αυτή η προσέγγιση, η έκφραση του μοναδικού γονιδίου αναφοράς πρέπει να παραμείνει σταθερή μεταξύ όλων των πειραματικών δειγμάτων. Για να βρεθεί ένας τέτοιος βολικός στόχος απαιτείται πρόσθετη επικύρωση, ωστόσο ακόμη και σήμερα η πολύ κοινή και λανθασμένη προσέγγιση είναι να επιλέγουμε αυτό το γονίδιο τυχαία χωρίς επικύρωση. Τα GAPDH, b actin και 18S είναι ιδιαίτερα αγαπημένα στη δημοσιευμένη βιβλιογραφία, που συνήθως χρησιμοποιούνται χωρίς επικύρωση ή αιτιολόγηση. Όταν το γονίδιο αναφοράς δεν εκφράζεται σταθερά μεταξύ όλων των δειγμάτων, μια αναλογία του γονιδίου στόχου που ενδιαφέρει προς το γονίδιο αναφοράς θα αντανακλά τις αλλαγές έκφρασης και των δύο στόχων. Αυτό δεν είναι χρήσιμο όταν η συμπεριφορά έκφρασης κανενός στόχου δεν έχει οριστεί και μπορεί να οδηγήσει σε ανακρίβειες και ψευδή αποτελέσματα7. Μια τροποποίηση αυτής της προσέγγισης είναι η επικύρωση του επιλεγμένου γονιδίου αναφοράς8 ή η επιλογή ενός σημαντικού γονιδίου αναφοράς με καθορισμένη βιολογία όπου η μεταγραφική απόκριση είναι καλά χαρακτηρισμένη (μια προσέγγιση που αναφέρεται ως κανονικοποίηση σε Ειδική Εσωτερική Αναφορά ή SIR). Με αυτόν τον τρόπο η βιολογία του γονιδίου στόχου εκφράζεται σε σχέση με την αλλαγή στη βιολογία του γονιδίου αναφοράς. Το πρόβλημα με τη χρήση αυτών των προσεγγίσεων μεμονωμένου γονιδίου αναφοράς είναι ότι η ανάλυσή τους (ελάχιστη μέτρηση αυτοπεποίθησης) περιορίζεται στο ελάχιστο σφάλμα της τεχνικής.

An alternative approach, which is the current gold standard, is to express data on the behavior of multiple reference genes and use geometric averaging to measure and control for trends caused by errors in relative reference gene expression. There are several approaches such as those offered by geNorm or Normfinder10,11 that allow comparing the amount of potential reference genes in selected groups of experimental and control samples. They use the combined variation of several genes to produce a stable reference factor for normalization. However, the requirement for multiple reference transcript measurements, although highly accurate and potentially providing 0.5-fold resolution12, is time, sample and resource intensive. In addition, the use of a single reporter gene, SIR, or multiple reporter genes requires appropriate validation. Recent developments point to a possible solution to this conundrum. This method normalizes the expression of a target gene of interest in several hundred different transcripts by targeting their integrated expressed ALU repeats (EAR) in primate systems or expressed B-element repeats (in mouse models). This avoids problems associated with any bias caused by the use of several genes, enables normalization that is no longer dependent on tissue or treatment, and promises to increase transparency and reproducibility of data. A similar approach can be used for microRNA expression normalization, where the mean expression value of all expressed miRNAs in a given sample is used as a normalization factor for real-time quantitative microRNA PCR data and appears to outperform current normalization strategies in terms of better reduction of technical variation and more accurate estimation biological changes13.

conclusions

Generating qPCR data is deceptively simple, but once mastered, the next challenge is to analyze the data so that they reflect underlying biological or clinical phenomena rather than technical flaws. Proper data analysis and normalization strategies are key aspects of the qPCR data analysis workflow, but are easily derailed by inappropriate approaches to the process. To reduce the risk of misinterpretation of qPCR data, Sigma Aldrich recently purchased GenEx software to provide greater support to scientists during the data analysis process14. There are also commercial companies, such as MultiD6 or Biogazelle15, that provide a qPCR data analysis service using their own software, GenEx or qbaseplus. The use of such software facilitates proper data analysis, resulting in better quality publications and contributing to, rather than confounding, scientific progress.

bibliographical references

1. WITH. Bustin, V. Beneš, J.A. Garson, J. Hellemans, J. Huggett, Μ. Cubist, R. Mueller, T. Nolan, M.W. Pfaffl, G.F. Shipley, J. Vandesompele and C.T. Wittwer, MIQE Guidelines: Minimum Information for Publishing Quantitative Real-Time PCR Experiments. Clinical Chemistry 55 (2009) 609-620.
2. I. Garson, J.F. Huggett, S.A. Bustin, M.W. Pfaffl, V. Benes, J. Vandesompele and G.L. Shipley, Unreliable real-time PCR analysis of human endogenous retrovirus-W (HERV-W) expression and DNA copy number in multiple sclerosis. AIDS Res. Hum. Retroviruses 25 (2009) 377-378.
3. WITH. Bustin and T. Nolan, Pitfalls of Real-Time Quantitative Reverse Transcription Polymerase Chain Reaction. J Biomol Tech 15 (2004) 155-66.
4. Tanya Nolan and Stephen A Bastin. Optimization of the PCR step of the qPCR assay. Eur Pharm Rev. 4 (2009) 15-20.
5. Huggett JF, Novak T, Garson JA, Green C, Morris-Jones SD, Miller RF, Zumla A. Differential sensitivity of PCR reactions to inhibitors: an important and unrecognized phenomenon. Notes BMC Res. Aug 28, 2008, 1:70 am.
6. http://www.multid.se/genex.html
7. Dheda K, Huggett JF, Chang JS, Kim LU, Bustin SA, Johnson MA, Rook GA, Zumla A. Consequences of using an inappropriate reference gene to normalize real-time reverse transcription-PCR data. Anal biochemistry. 2005 September 1; 344 (1): 141-3.
8. Dheda K, Huggett JF, Bustin SA, Johnson MA, Rook G, Zumla A. Validation of housekeeping genes for normalization of RNA expression by real-time PCR. Industrial. 2004 July;37(1):112-4, 116, 118-9.
9. S.A. Bustin & J.F. Huggett (unpublished)
10. Vandesompele J, De Preter K, Pattyn F, Poppe B, Van Roy N, De Paepe A, Speleman F. Accurate normalization of real-time quantitative RT-PCR data using the geometric average of multiple internal control genes. Genome Biol. 2002 June 18 3 (7)
11. Andersen CL, Jensen JL, Ørntoft TF. Normalization of real-time quantitative reverse transcription-PCR data: a model-based variance estimation approach to identify genes suitable for normalization applied to bladder and colon cancer datasets. Cancer Res. 2004 Aug 1;64(15):5245-50.
12. J. Hellemans, O. Preobraženska, A. Willaert, P. Debeer, P.C. Verdonk, T. Costa, K. Janssens, B. Menten, N. Van Roy, S.J. Vermeulen, R. Savarirayan, W. Van Hul, F. Vanhoenacker, D. Huylebroeck, A. De Paepe, J.M. Naeyaert, J. Vandesompele, F. Speleman, K. Verschueren, P.J. Coucke i G.R. Mortiera, mutacije gubitka funkcije u LEMD3 rezultiraju osteopoezom, Buschke-Ollendorffovim sindromom i melorheostozom. Nat Genet 36 (2004) 1213-8.
13. P. Mestdagh, P. Van Vlierberghe, A. De Weer, D. Muth, F. Westermann, F. Speleman and J. Vandesompele, A new and universal method for normalizing RT-qPCR microRNA data. Genome Biol 10 (2009) R64.
14. For more information email[emailprotected]
15. http://www.biogazelle.com