Back to Genetics     Back to Evolution     Back to HLA     Back to MHC     Back to Biostatistics     Glossary     Homepage

REAL-TIME PCR

M.Tevfik Dorak, MD, PhD

Mirror Site  (no pop-up ads)

Real-time reverse-transcriptase (RT) PCR quantitates the initial amount of the template most specifically, sensitively and reproducibly, and is a preferable alternative to other forms of quantitative RT-PCR, which detects the amount of final amplified product. Real-time PCR monitors the fluorescence emitted during the reaction as an indicator of amplicon production during each PCR cycle (i.e., in real time) as opposed to the endpoint detection by conventional quantitative PCR methods. The real-time progress of the reaction can be viewed in some systems (such as Cepheid). The real-time RT-PCR does not detect the size of the amplicon and thus does not allow the differentiation between DNA and cDNA amplification, however, it is not influenced by non-specific amplification unless SYBR Green is used (see below). Real-time PCR quantitation eliminates post-PCR processing of PCR products (which is necessary in competitive RT-PCR). This helps to increase throughput, reduce the chances of carryover contamination and remove post-PCR processing as a potential source of error. In comparison to conventional RT-PCR, real-time PCR also offers a much wider dynamic range of up to 10⁷-fold (compared to 1000-fold in conventional RT-PCR). This means that a wide range of ratios of target and normalizer can be assayed with equal sensitivity and specificity. It follows that the broader the dynamic range, the more accurate the quantitation.

The real-time PCR system is based on the detection and quantitation of a fluorescent reporter. This signal increases in direct proportion to the amount of PCR product in a reaction. By recording the amount of fluorescence emission at each cycle, it is possible to monitor the PCR reaction during exponential phase where the first significant increase in the amount of PCR product correlates to the initial amount of target template. There are two general methods for the quantitative detection of the amplicon: (1) fluorescent probes or (2) DNA-binding agents. The TaqMan probes (Heid et al., 1996) and molecular beacons (and more recently, scorpions) use the fluorogenic 5' exonuclease activity of Taq polymerase to measure the amount of target sequences in cDNA samples. TaqMan probes are oligonucleotides longer than the primers (20-30 bases long with a Tm value of 10 ^oC higher) that contain a fluorescent dye usually on the 5' base, and a quenching dye (usually TAMRA) typically on the 3' base. When irradiated, the excited fluorescent dye transfers energy to the nearby quenching dye molecule rather than fluorescing (this is called FRET = Förster or fluorescence resonance energy transfer). Thus, the close proximity of the reporter and quencher prevents emission of any fluorescence while the probe is intact. TaqMan probes are designed to anneal to an internal region of a PCR product. When the polymerase replicates a template on which a TaqMan probe is bound, its 5' exonuclease activity cleaves the probe. This ends the activity of quencher (no FRET) and the reporter dye starts to emit fluorescence which increases in each cycle proportional to the rate of probe cleavage. Accumulation of PCR products is detected by monitoring the increase in fluorescence of the reporter dye (note that primers are not labeled). TaqMan assay uses universal thermal cycling parameters and PCR reaction conditions. Because the cleavage occurs only if the probe hybridizes to the target, the fluorescence detected originates from specific amplification. The process of hybridization and cleavage does not interfere with the exponential accumulation of the product. One specific requirement for fluorogenic probes is that there be no G at the 5' end. A 'G' adjacent to the reporter dye quenches reporter fluorescence even after cleavage.

Molecular beacons also contain fluorescent (FAM, TAMRA, TET, ROX) and quenching dyes (typically DABCYL) at either end but they are designed to adopt a hairpin structure while free in solution to bring the fluorescent dye and the quencher in close proximity for FRET to occur. It is designed to have two arms with complementary sequences that form a very stable hybrid or stem. The close proximity of the reporter and the quencher in this hairpin configuration suppresses reporter fluorescence. When the beacon hybridizes to the target during the annealing step, the reporter dye is separated from the quencher and the reporter fluoresces (FRET does not occur). Molecular beacons remain intact during PCR and must rebind to target every cycle for fluorescence emission. This will correlate to the amount of PCR product available. Both TaqMan probes and molecular beacons allow detection of multiple DNA species (multiplexing) by use of different reporter dyes on different probes/beacons. By multiplexing, both the target and endogenous control can be amplified in single tube.

The cheaper alternative is the double-stranded DNA binding dye chemistry, which quantitates the amplicon production (including non-specific amplification and primer-dimer complex) by the use of a non-sequence specific fluorescent intercalating agent (SYBR-green I or ethidium bromide). SYBR Green I is a minor groove binding dye. It does not bind to ssDNA. The major problem with SYBR Green-based detection is that non-specific amplifications cannot be distinguished from specific amplifications. A relatively minor and more controllable problem is that longer amplicons create a stronger signal (if combined with other factors, this may cause CDC camera saturation, see below). Obviously SYBR green can only be used in singleplex reactions.

The threshold cycle or the C_T value is the cycle at which a significant increase in DRn is first detected (for definition of DRn, see below). The threshold cycle is when the system begins to detect the increase in the signal associated with an exponential growth of PCR product during the log-linear phase. This phase provides the most useful information about the reaction (certainly more important than the end point). The slope of the log-linear phase is a reflection of the amplification efficiency. For the slope to be an indicator of real amplification (rather than signal drift), there has to be an inflection point. This is the point on the growth curve when the log-linear phase begins. It also represents the greatest rate of change along the growth curve. (Signal drift is characterized by gradual increase or decrease in fluorescence without amplification of the product.) The important parameter for quantitation is the C_T. The higher the initial amount of genomic DNA, the sooner accumulated product is detected in the PCR process, and the lower the C_T value. The choice of threshold, which will determine the C_T value is up to the operator and one of the subjective elements in real-time PCR. It should be placed above any baseline activity and within the exponential increase phase (which looks linear in the log transformation). The SmartCycler software allows determination of the cycle threshold by a mathematical analysis of the growth curve. This provides better run-to-run reproducibility. A C_T value of 40 means no amplification and this value cannot be included in the calculations. Besides being used for quantitation, the C_T value can be used for qualitative analysis as a pass/fail measure.

Relative gene expression comparisons work best when the gene expression of the chosen endogenous control is more abundant and remains constant, in proportion to total RNA, among the samples. By using an endogenous control as an active reference, quantitation of an mRNA target can be normalized for differences in the amount of total RNA added to each reaction. For this purpose, the most common choices are 18S RNA, GAPDH (glyceraldehyde-3-phosphate dehydrogenase) and b-actin. Because the 18S mRNA does not have a poly-A tail, cDNA synthesis using oligo-dT should not be used if 18S RNA will be used as a normalizer. The issue of the choice of a normalizer has recently been reviewed by Suzuki et al. (BioTechniques 2000;29:332). The authors recommend caution in the use of GAPDH as a normalizer as it has been shown that its expression may be upregulated in proliferating cells. They recommend b-actin as a better active reference. GAPDH is also severely criticized as a normalizer in another recent review (Bustin SA, 2000). Caution should also be exercised when 18S RNA is used as a normalizer as it is a ribosomal RNA species (not mRNA) and may not always represent the overall cellular mRNA population. Since the chosen mRNA species should be proportional to the amount of input RNA, it may be best to use a combination as normalizer. It is desirable to validate the chosen normalizer for the target cell or tissue. It should be expressed at a constant level at different time points by the same individual and also by different individuals at the target cell or tissue (for example, peripheral blood lymphocytes). This aim can be achieved by the ABI's TaqMan Human Endogenous Control Plate which evaluates the expression of eleven select housekeeping genes. Our own experience showed that b-actin or 18S RNA are the best choices as normalizers for the peripheral blood mononuclear cells, whereas GAPDH performed worst.

Multiplex TaqMan assays can be performed using multiple dyes with distinct emission wavelengths. Available dyes for this purpose are FAM, TET, VIC and JOE (the most expensive). TAMRA is reserved as the quencher on the probe and ROX as the passive reference. For best results, the combination of FAM (target) and VIC (endogenous control) is recommended (they have the largest difference in emission maximum) whereas JOE and VIC should not be combined. It is important that if the dye layer has not been chosen correctly, the machine will still read the other dye's spectrum. For example, both VIC and FAM emit fluorescence in a similar range to each other and when doing a single dye, the wells should be labeled correctly. In the case of multiplexing, the spectral compensation for the post run analysis should be turned on (on ABI 7700: Instrument/Diagnostics/Advanced Options/Miscellaneous). Activating spectral compensation improves dye spectral resolution.

The one-step real-time RT-PCR performs reverse transcription and PCR in a single buffer system and in one tube. In two-step RT-PCR, these two steps are performed separately in different tubes. For multiplex real-time RT-PCR, one-step PCR cannot be used.

TaqMan primer and probe design guidelines:

1. The Primer Express software designs primers with a melting temperature (Tm) of 58-60⁰ C, and probes with a Tm value of 10⁰ C higher. The Tm of both primers should be equal,

2. Primers should be 15-30 bases in length,

3. The G+C content should ideally be 30-80%. If a higher G+C content is unavoidable, the use of high annealing and melting temperatures, cosolvents such as glycerol, DMSO, or 7-deaza-dGTP may be necessary,

4. The run of an identical nucleotide should be avoided. This is especially true for G, where runs of four or more Gs is not allowed,

5. The total number of Gs and Cs in the last five nucleotides at the 3' end of the primer should not exceed two (the newer version of the software has an option to do this automatically). This helps to introduce relative instability to the 3' end of primers to reduce non-specific priming. The primer conditions are the same for SYBR Green assays,

6. Maximum amplicon size should not exceed 400 bp (ideally 50-150 bases). Smaller amplicons give more consistent results because PCR is more efficient and more tolerant of reaction conditions (the short length requirement has nothing to do with the efficiency of 5' nuclease activity),

7. The probes should not have runs of identical nucleotides (especially four or more consecutive Gs), G+C content should be 30-80%, there should be more Cs than Gs, and not a G at the 5' end. The higher number of Cs produces a higher DRn. The choice of probe should be made first,

8. To avoid false-positive results due to amplification of contaminating genomic DNA in the cDNA preparation, it is preferable to have primers spanning exon-exon junctions. This way, genomic DNA will not be amplified (the PDAR kit for human GAPDH amplification has such primers),

9. If a TaqMan probe is designed for allelic discrimination, the mismatching nucleotide (the polymorphic site) should be in the middle of the probe rather than at the ends,

10. Use primers that contain dA nucleotides near the 3' ends so that any primer-dimer generated is efficiently degraded by AmpErase UNG (mentioned in p.9 of the manual for EZ RT-PCR kit; P/N 402877). If primers cannot be selected with dA nucleotides near the ends, the use of primers with 3' terminal dU-nucleotides should be considered.

(See also the general principles of PCR Primer Design by InVitroGen.)

General recommendations for real-time RT-PCR:

1. Use positive-displacement pipettes to avoid inaccuracies in pipetting,

2. The sensitivity of real-time PCR allows detection of the target in 2 pg of total RNA. The number of copies of total RNA used in the reaction should ideally be enough to give a signal by 25-30 cycles (preferably less than 100 ng). The amount used should be decreased or increased to achieve this,

3. The optimal concentrations of the reagents are as follows:

i. Magnesium chloride concentration should be between 4 and 7 mM. It is optimized as 5.5 mM for the primers/probes designed using the Primer Express software,

ii. Concentrations of dNTPs should be balanced with the exception of dUTP (if used). Substitution of dUTP for dTTP for control of PCR product carryover requires twice dUTP that of other dNTPs. While the optimal range for dNTPs is 500 mM to 1 mM (for one-step RT-PCR), for a typical TaqMan reaction (PCR only), 200 mM of each dNTP (400 mM of dUTP) is used,

iii. Typically 0.25 mL (1.25 U) AmpliTaq DNA Polymerase (5.0 U/mL) is added into each 50 mL reaction. This is the minimum requirement. If necessary, optimization can be done by increasing this amount by 0.25 U increments,

iv. The optimal probe concentration is 50-200 nM, and the primer concentration is 100-900 nM. Ideally, each primer pair should be optimized at three different temperatures (58, 60 and 62⁰ C for TaqMan primers) and at each combination of three concentrations (50, 300, 900 nM). This means setting up three different sets (for three temperatures) with nine reactions in each (50/50 mM, 50/300 mM, 50/900, 300/50, 300/300, 300/900, 900/50, 900/300, 900/900 mM) using a fixed amount of target template. If necessary, a second round of optimization may improve the results. Optimal performance is achieved by selecting the primer concentrations that provide the lowest C_T and highest DRn. Similarly, the probe concentration should be optimized for 25-225 nM,

4. If AmpliTaq Gold DNA Polymerase is being used, there has to be a 9-12 min pre-PCR heat step at 92 - 95⁰ C to activate it. If AmpliTaq Gold DNA Polymerase is used, there is no need to set up the reaction on ice. A typical TaqMan reaction consists of 2 min at 50⁰ C for UNG (see below) incubation, 10 min at 95⁰ C for Polymerase activation, and 40 cycles of 15 sec at 95⁰ C (denaturation) and 1 min at 60⁰ C (annealing and extension). A typical reverse transcription cycle (for cDNA synthesis), which should precede the TaqMan reaction if the starting material is total RNA, consists of 10 min at 25⁰ C (primer incubation), 30 min at 48⁰ C (reverse transcription with conventional reverse transcriptase) and 5 min at 95⁰ C (reverse transcriptase inactivation),

5. AmpErase uracil-N-glycosylase (UNG) is added in the reaction to prevent the reamplification of carry-over PCR products by removing any uracil incorporated into amplicons. This is why dUTP is used rather than dTTP in PCR reaction. UNG does not function above 55 ⁰C and does not cut single-stranded DNA with terminal dU nucleotides. UNG-containing master mix should not be used with one-step RT-PCR unless rTth DNA polymerase is being used for reverse transcription and PCR (TaqMan EZ RT-PCR kit),

6. It is necessary to include at least three No Amplification Controls (NAC) as well as three No Template Controls (NTC) in each reaction plate (to achieve a 99.7% confidence level in the definition of +/- thresholds for the target amplification, six replicates of NTCs must be run). NAC former contains sample and no enzyme. It is necessary to rule out the presence of fluorescence contaminants in the sample or in the heat block of the thermal cycler (these would cause false positives). If the absolute fluorescence of the NAC is greater than that of the NTC after PCR, fluorescent contaminants may be present in the sample or in the heating block of the thermal cycler,

7. The dynamic range of a primer/probe system and its normalizer should be examined if the DDC_T method is going to be used for relative quantitation. This is done by running (in triplicate) reactions of five RNA concentrations (for example, 0, 80 pg/mL, 400 pg/mL, 2 ng/mL and 50 ng/mL). The resulting plot of log of the initial amount vs C_T values (standard curve) should be a (near) straight line for both the target and normalizer real-time RT-PCRs for the same range of total RNA concentrations,

8. The passive reference is a dye (ROX) included in the reaction (present in the TaqMan universal PCR master mix). It does not participate in the 5' nuclease reaction. It provides an internal reference for background fluorescence emission. This is used to normalize the reporter-dye signal. This normalization is for non-PCR-related fluorescence fluctuations occurring well-to-well (concentration or volume differences) or over time and different from the normalization for the amount of cDNA or efficiency of the PCR. Normalization is achieved by dividing the emission intensity of reporter dye by the emission intensity of the passive reference. This gives the ratio defined as Rn,

9. If multiplexing is done, the more abundant of the targets will use up all the ingredients of the reaction before the other target gets a chance to amplify. To avoid this, the primer concentrations for the more abundant target should be limited.

Recommendations for the general assay of cDNA samples:

1. Reverse transcription of total RNA to cDNA should be done with random hexamers (not with oligo-dT). If oligo-dT has to be used long mRNA transcripts or amplicons greater than two kilobases upstream should be avoided, and 18S RNA cannot be used as normalizer,

2. Multiplex PCR will only work properly if the control primers are limiting (ABI control reagents do not have their primers limited),

3. The range of target cDNA used is 10 ng to 1 mg. If DNA is used (mainly for allelic discrimination studies), the optimum amount is 100 ng to 1 mg,

4. It is ideal to treat each RNA preparation with RNAse free DNAse to avoid genomic DNA contamination. Even the best RNA extraction methods yield some genomic DNA. Of course, it is ideal to have primers not amplifying genomic DNA at all but sometimes this may not be possible,

5. For optimal results, the reagents (before the preparation of the PCR mix) and the PCR mixture itself (before loading) should be vortexed and mixed well. Otherwise there may be shifting Rn value during the early (0 - 5) cycles of PCR. It is also important to add probe to the buffer component and allow it to equilibrate at room temperature prior to reagent mix formulation.

TaqMan primers and probes:

The TaqMan probes ordered from ABI at midi-scale arrive already resuspended at 100 mM. If a 1/20 dilution is made, this gives a 5 mM solution. This stock solution should be aliquoted, frozen and kept in the dark. Using 1 mL of this in a 50 mL reaction gives the recommended 100 nM final concentration.

The primers arrive lyophilized with the amount given on the tube in pmols (such as 150.000 pmol which is equal to 150 nmol). If X nmol of primer is resuspended in X mL of H₂O, the resulting solution is 1 mM. It is best to freeze this stock solution in aliquots. When the 1 mM stock solution is diluted 1/100, the resulting working solution will be 10 mM. To get the recommended 50 - 900 nM final primer concentration in 50 mL reaction volume, 0.25 - 4.50 mL should be used per reaction (2.5 mL for 500 nM final concentration).

The PDAR primers and probes are supplied as a mix in one tube. They have to be used 2.5 mL in a 50 mL reaction volume.

Setting up one-step TaqMan reaction:

One-step real-time PCR uses RNA (as opposed to cDNA) as a template. This is the preferred method if the RNA solution has a low concentration but only if singleplex reactions are run. The disadvantage is that RNA carryover prevention enzyme AmpErase cannot be used in one-step reaction format. In this method, both reverse transcriptase and real-time PCR take place in the same tube. The downstream PCR primer also acts as the primer for reverse transcriptase (random hexamers or oligo-dT cannot be used for reverse transcription in one-step RT-PCR). One-step reaction requires higher dNTP concentration (³ 300 mM vs 200 mM) as it combines two reactions needing dNTPs in one. A typical reaction mix for one-step PCR by Gold RT-PCR kit is as follows:

H₂O + RNA : 20.5 mL [24 mL if PDAR is used]

10X TaqMan buffer : 5.0 mL

MgCl₂ (25 mM) : 11.0 mL

dATP (10mM) : 1.5 mL [for final concentration of 300 mM]

dCTP (10mM) : 1.5 mL [for final concentration of 300 mM]

dGTP (10mM) : 1.5 mL [for final concentration of 300 mM]

dUTP (20mM) : 1.5 mL [for final concentration of 600 mM]

Primer F (10 mM) * : 2.5 mL [for final concentration of 500 nM]

Primer R (10 mM) * : 2.5 mL [for final concentration of 500 nM]

TaqMan Probe * : 1.0 mL [for final concentration of 100 nM]

AmpliTaq Gold : 0.25 mL [can be increased for higher efficiency]

Reverse Transcriptase : 0.25 mL

RNAse inhibitor : 1.00 mL

* If a PDAR is used, 2.5 mL of primer + probe mix used.

Ideally 10 pg - 100 ng RNA should be used in this reaction. Note that decreasing the amount of template from 100 ng to 50 ng will increase the C_T value by 1. To decrease a C_T value by 3, the initial amount of template should be increased 8-fold. ABI claims that 2 picogram RNA can be detected by this system and the maximum amount of RNA that can be used is 1 microgram. For routine analysis, 10 pg - 100 ng RNA and 100 pg - 1 mg genomic DNA can be used.

Cycling parameters for one-step PCR:

Reverse transcription (by MuLV) 48⁰ C for 30 min

AmpliTaq activation 95⁰ C for 10 min

PCR: denaturation 95⁰ C for 15 sec and annealing/extension 60⁰ C for 1 min  (repeated 40 times)

(On ABI 7700, minimum holding time is 15 seconds.)

The recently introduced EZ one-step RT-PCR kit allows the use of UNG as the incubation time for reverse transcription is 60 ⁰C thanks to the use of a thermostable reverse transcriptase. This temperature also a better option to avoid primer dimers and non-specific bindings at 48 ⁰C.

Operating ABI 7700:

Make sure the following before starting a run:

1. Cycle parameters are correct for the run (somebody may have used different parameters before you),

2. Choice of spectral compensation is correct (off for singleplex, on for multiplex reactions),

3. Choice of "Number of PCR Stages" is correct in the Analysis Options box (Analysis/Options). This may have to be manually assigned after a run if the data is absent in the amplification plot but visible in the plate view, and the X-axis of the amplification is displaying a range of 0-1 cycles,

4. No Template Control is labeled as such (for accurate DRn calculations),

5. The choice of dye component should be made correctly before data analysis. Even if the probe is labeled with FAM and VIC is chosen there will be some result but the wrong one,

6. You must save the run before it starts by giving it a name (not leaving as untitled). Also at the end of the run, first save the data before starting to analyze,

7. The ABI software requires extreme caution. Do not attempt to stop a run after clicking on the Run button. You will have problems and if you need to switch off and on the machine, you have to wait for at least an hour to restart the run.

When analyzing the data remember that the default setting for baseline is 3 - 15. If any C_T value is <15, the baseline should be changed accordingly (the baseline stop value should be 1-2 smaller than the smallest C_T value). For a useful discussion of this matter, see the ABI Tutorial on Setting Baselines and Thresholds. (Interestingly, this issue is best discussed in the manual for TaqMan Human Endogenous Control Plate.)

If the results do not make sense, check the raw spectra for a possible CDC camera saturation during the run. Saturation of CDC camera may be prevented by using optical caps rather than optical adhesive cover. It is also more likely to happen when SYBR Green I is used, when multiplexing and when a high concentration of probe is used.

Interpretation of results:

At the end of each reaction, the recorded fluorescence intensity is used for the following calculations:

Rn⁺ is the Rn value of a reaction containing all components, Rn^- is the Rn value of an unreacted sample (baseline value or the value detected in NTC). DRn is the difference between Rn⁺ and Rn^-. It is an indicator of the magnitude of the signal generated by the PCR.

There are three methods to quantitate the amount of template:

1. Absolute standard method: In this method, a known amount of standard such as in vitro translated RNA (cRNA) is used,

2. Relative standard: Known amounts of the target nucleic acid are included in the assay design in each run,

3. Comparative C_T method: This method uses no known amount of standard but compares the relative amount of the target sequence to any of the reference values chosen and the result is given as relative to the reference value (such as the expression level of resting lymphocytes or a standard cell line).

The comparative C_T method (DDC_T) for relative quantitation of gene expression:

This method enables relative quantitation of template and increases sample throughput by eliminating the need for standard curves when looking at expression levels relative to an active reference control (normalizer). For this method to be successful, the dynamic range of both the target and reference should be similar. A sensitive method to control this is to look at how DC_T (the difference between the two CT values of two PCRs for the same initial template amount) varies with template dilution. If the efficiencies of the two amplicons are approximately equal, the plot of log input amount versus DC_T will have a nearly horizontal line (a slope of <0.10). This means that both PCRs perform equally efficiently across the range of initial template amounts. If the plot shows unequal efficiency, the standard curve method should be used for quantitation of gene expression. The dynamic range should be determined for both (1) minimum and maximum concentrations of the targets for which the results are accurate and (2) minimum and maximum ratios of two gene quantities for which the results are accurate. In conventional competitive RT-PCR, the dynamic range is limited to a target-to-competitor ratio of about 10:1 to 1:10 (the best accuracy is obtained for 1:1 ratio). The real-time PCR is able to achieve a much wider dynamic range.

Running the target and endogenous control amplifications in separate tubes and using the standard curve method requires the least amount of optimization and validation. The advantage of using the comparative C_T method is that the need for a standard curve is eliminated (more wells are available for samples). It also eliminates the adverse effect of any dilution errors made in creating the standard curve samples.

As long as the target and normalizer have similar dynamic ranges, the comparative C_T method (DDC_T method) is the most practical method. It is expected that the normalizer will have a higher expression level than the target (thus, a smaller C_T value). The calculations for the quantitation start with getting the difference (DC_T) between the C_T values of the target and the normalizer:

DC_T = C_T (target) - C_T (normalizer)

This value is calculated for each sample to be quantitated (unless, the target is expressed at a higher level than the normalizer, this should be a positive value. It is no harm if it is negative). One of these samples should be chosen as the reference (baseline) for each comparison to be made. The comparative DDC_T calculation involves finding the difference between each sample's DC_T and the baseline's DC_T. If the baseline value is representing the minimum level of expression, the DDC_T values are expected to be negative (because the DC_T for the baseline sample will be the largest as it will have the greatest C_T value). If the expression is increased in some samples and decreased in others, the DDC_T values will be a mixture of negative and positive ones. The last step in quantitation is to transform these values to absolute values. The formula for this is:

comparative expression level = 2 ^{- DDCT}

For expressions increased compared to the baseline level this will be something like 2³ = 8 times increase, and for decreased expression it will be something like 2^-3 = 1/8 of the reference level. Microsoft Excel can be used to do these calculations by simply entering the C_T values (there is an online ABI tutorial on the use of spread sheet programs to produce amplification plots; the TaqMan Human Endogenous Control Plate protocol also contains detailed instructions on using MS Excel for real-time PCR data analysis).

The other (absolute) quantification methods are outlined in the ABI User Bulletins. The Bulletins #2 and #5 are most useful for the