Biomolecular Detection And Quantification
Biomolecular Detection and Quantification 9 (2016) 29–39Contents lists available at ScienceDirectBiomolecular Detection and Quantificationjournal homepage: www.elsevier.com/locate/bdqResearch paperValidation of a digital PCR method for quantification of DNA copynumber concentrations by using a certified reference materialLiesbet Deprez , Philippe Corbisier, Anne-Marie Kortekaas, Stéphane Mazoua,Roxana Beaz Hidalgo, Stefanie Trapmann, Hendrik EmonsDirectorate for Health, Consumers and Reference Materials, Joint Research Centre, European Commission, Retieseweg 111, 2440 Geel, Belgiuma r t i c l ei n f oArticle history:Received 30 June 2016Received in revised form 9 August 2016Accepted 19 August 2016Available online 30 August 2016Handled by Justin O’GradyKeywords:Digital PCRMethod validationMeasurement uncertaintyCertified reference materialsa b s t r a c tDigital PCR has become the emerging technique for the sequence-specific detection and quantificationof nucleic acids for various applications. During the past years, numerous reports on the developmentof new digital PCR methods have been published. Maturation of these developments into reliable analytical methods suitable for diagnostic or other routine testing purposes requires their validation for theintended use.Here, the results of an in-house validation of a droplet digital PCR method are presented. This methodis intended for the quantification of the absolute copy number concentration of a purified linearizedplasmid in solution with a nucleic acid background. It has been investigated which factors within themeasurement process have a significant effect on the measurement results, and the contribution to theoverall measurement uncertainty has been estimated. A comprehensive overview is provided on all theaspects that should be investigated when performing an in-house method validation of a digital PCRmethod. 2016 The Authors. Published by Elsevier GmbH. This is an open access article under the CC BYlicense (http://creativecommons.org/licenses/by/4.0/).1. IntroductionAccurate quantification of the copy number concentration ofspecific nucleic acid sequences is important for several applicationsboth within the fields of red biotechnology, (e.g. oncology and infectious diseases) and green biotechnology (e.g. GMO testing). Duringthe last decade, digital PCR (dPCR) has shown to be the emergingtechnique for the sequence-specific detection and quantificationof nucleic acids [1,2]. The measurement principle of dPCR relies onpartitioning the PCR mix across a large number of small individualreaction volumes, such that the distribution of the target sequencefollows a binominal distribution function and that a part of thereaction volumes does not contain a copy of the target sequence[3]. Following an end-point PCR, partitions containing one or morecopies of the target sequence are labelled positive and counted. Theproportion of positive partitions is used to estimate the copy number concentration of the target sequence, taking into account thestatistics of the binominal distribution [4]. Commercially availabledPCR systems are based on two different approaches to partitionthe PCR mix: some use microfluidic chips on which the PCR mix is Corresponding author.E-mail address: Liesbet.deprez@ec.europa.eu (L. Deprez).distributed over premanufactured chambers [5,6] while others arebased on oil-water emulsions to separate the solution into droplets[7,8].Digital PCR has the potential to replace quantitative real-timePCR (qPCR) for several of the current applications as it can haveseveral advantages, including improved precision [9], reducedinterference of PCR inhibitors [10] and independence of a calibration curve to determine the copy number concentration of thetarget sequence [11]. However, the measurement principle of thedPCR implies some essential prerequisites and failure to fulfil one ormore of these, affects the reliability of the measured absolute copynumber concentrations. First, the copies of the target sequenceshould be distributed over the partitions in a random and uniform manner meaning that there should be no aggregation of DNAsequences. Second, the volume of the partitions should be wellknown and consistent within and between measurements. Third,partitions should be correctly classified as positive or negative afterthe end-point PCR [12].Numerous reports on the development of new dPCR methodshave been published during the past years. Maturation of these newdevelopments into reliable analytical methods suitable for diagnostic or other routine testing purposes requires that the methodsare validated for their intended use. Method validation is the toolto proof that a method is fit for purpose and to ensure that -7535/ 2016 The Authors. Published by Elsevier GmbH. This is an open access article under the CC BY license (http://creativecommons.org/licenses/by/4.0/).
30L. Deprez et al. / Biomolecular Detection and Quantification 9 (2016) 29–39Table 1Critical performance characteristics which should be assessed during the validationof a quantitative analytical method.Performance characteristicDescriptionSelectivityDegree to which the method canquantify the particular analyte (i.e. aspecific target sequence) accurately inthe presence of interfering substanceswhich could be present in the samples.The analyte concentration interval overwhich the method provides resultswith an acceptable uncertainty. In thisconcentration range, the relationshipbetween response and concentration iscontinuous, reproducible and linearafter suitable data transformation.The closeness of agreement between ameasurement result produced by themethod for the analyte in a certainsample and the accepted referencevalue of that analyte. Accuracy can bedivided into two parts:Measure of the variability inindependent measurement resultsobtained for the same sample understipulated conditions. There are threedifferent levels depending on theconditions: repeatability, intermediateprecision and reproducibility.The closeness of agreement betweenthe mean of an infinite number ofmeasurement results produced by themethod for the analyte in a certainsample and the accepted referencevalue of that analyte.Interval associated with ameasurement result which expressesthe range of values that can reasonablybe attributed to the analyte beingmeasured.The lowest analyte concentration thatcan be distinguished from zero, with aspecified level of confidence.The lowest analyte concentration forwhich the method provides resultswith an acceptable uncertainty.Measure of the capacity of the methodto remain unaffected by small, butdeliberate variations in methodparameters.Working rangeAccuracy Precision TruenessMeasurement uncertaintyLimit of detection (LOD)Limit of quantification (LOQ)Robustness (or ruggedness)The descriptions given in this table are based on the definitions and explanations asprovided in several guidance documents [15–17].measurement results are sufficiently reliable so that related decisions can be taken with confidence. International standards suchas ISO/IEC 17025 [13] and ISO 15189 [14] also stress the needfor method validation. There are several guidance documents onmethod validation [15–17] describing a series of tests that both verify the assumptions on which the analytical method is based andestablish the performance characteristics of the method. Table 1provides a list of performance characteristics that are typicallyassessed during method validation. Several of these performancecharacteristics are also included in the guidelines on MinimumInformation for the publication of Quantitative dPCR Experiments(dMIQE) [18].Here, a complete in-house validation is described for a dPCRmethod using the droplet digitalTM PCR (ddPCR) system (Bio-Rad)which partitions the PCR mix in approximately 20,000 dropletswith an individual volume 1 nL. This ddPCR method amplifies aspecific sequence of the human fusion transcript BCR-ABL and isintended to be used for the quantification of the absolute copynumber concentration of a linearized plasmid carrying the BCRABL sequence in solution with a nucleic acid background. Theapproaches for method validation described in the following canbe used as an example for the validation of other dPCR methods.2. Materials and methods2.1. Test materialThe method validation was performed on samples of certifiedreference materials from the ERM-AD623 set [19]. ERM-AD623consists of 6 solutions of a double-stranded linearized plasmidcarrying 3 DNA fragments specific for 3 human cDNA transcripts:the transcript of the breakpoint cluster region gene (BCR), thetranscript of the glucuronidase beta gene (GUSB) and the aberranttranscript (BCR-ABL b3a2) consisting of a fusion of the BCR genewith the c-abl oncogene 1 (ABL). Each of the six solutions; ERMAD623a, ERM-AD623b, ERM-AD623c, ERM-AD623d, ERM-AD623eand ERM-AD623f has a different certified copy number concentration: (1.08 106 0.13 106 ), (1.08 105 0.11 105 ),(1.02 103 0.09 103 ),(1.03 104 0.10 104 ),(1.04 102 0.10 102 ) and (10.0 1.5) copies(cp)/ L, respectively. The plasmid solutions were prepared in a T1 E0.01 buffer(1 mM Tris, 0.01 mM EDTA, pH 8.0) supplemented with 50 mg/Lof transfer RNA from Escherichia coli (E. coli). The certified copynumber concentrations and the associated uncertainties assignedto the ERM-AD623 solutions were derived from measurementdata of 3 metrology institutes using a chip-based dPCR technology(i.e. the BioMarkTM system with 12.765 digital Arrays TM fromFluidigm).2.2. ddPCR methodThe ddPCR method validated in this study targets a sequencespecific for the human BCR-ABL transcript (referred here as theBCR-ABL ddPCR method). Also, a second ddPCR method was appliedtargeting a sequence specific for the ABL transcript (called the ABLddPCR method). These ddPCR methods are based on two qPCRmethods which were developed within the frame of a ‘EuropeAgainst Cancer’ program [20,21]. The sequences of the primersand probes and their concentrations used in the ddPCR methodscan be found in Supplementary data Table 1. The term ‘assay’ isused to refer to the combination of the specific primers and probes.All primers and probes were purified by HPLC (Life TechnologiesEurope BV). The PCR mix comprised 1 ddPCR Supermix for Probes(Bio-Rad, cat no. 186-3010), suitable primers and probes, nuclease free water (Promega, cat no. P1193) and the DNA sample. Tominimise the uncertainty from pipetting, all components, excluding the DNA sample, were premixed in the pre-sample mix, andthe final PCR mix was prepared gravimetrically by combining theDNA sample with the pre-sample mix using a microbalance. Thedensity of the pre-sample mix was determined by pipetting 100 Lon the microbalance using a calibrated pipette. The average density and the associated standard deviation (STD) of 10 replicatemeasurements were 1.0353 0.0026 g/L.Twenty microliters of the PCR mix were pipetted into the compartments of the Droplet Generator DG8TM Cartridge (Bio-Rad, 2types were used: cat no. 186-3008 and 186-4008) and 70 L ofthe Droplet Generation Oil for Probes (Bio-Rad, cat no. 186-3005)was added to the appropriate wells. The cartridges were coveredwith DG8TM Gaskets (Bio-Rad, cat no. 186-3009) and placed in aQX100TM Droplet Generator (Bio-Rad, cat no. 186-3002) to generate the droplets. Afterwards, the droplets were gently transferred toa semi-skirted and PCR clean 96-well PCR plate (Eppendorf, cat no.0030 128.605) using a Pipet-lite TM XLS manual 8-channel pipettewith the range 5–50 L (Rainin, cat no. L8-50XLS ). The PCR platewas sealed with pierceable foil (Bio-Rad, cat no. 181-4040) using
L. Deprez et al. / Biomolecular Detection and Quantification 9 (2016) 29–39a PX1TM PCR Plate Sealer (Bio-Rad, cat no. 181-4000). After sealing, the PCR plate was placed in a C1000 TouchTM Thermal Cycler(Bio-Rad, cat no. 185-1197) for PCR amplification. The PCR protocol can be found in the Supplementary data Table 2. The dropletreading was done with the QX 100 Droplet reader (Bio-Rad, catno. 186-3001) using ddPCRTM Droplet Reader Oil (Bio-Rad, cat no.186-3004).point between the average fluorescence amplitudes of the positiveand negative droplet cluster, no droplets in matrix blank replicateswere classified as positive (0/61275) and only 0.057 % (33/57895)of the droplets in the positive control replicates were classified asnegative.3.2. Working range2.3. Data analysisData acquisition and analysis were performed with the softwarepackage QuantaSoft (Bio-Rad). As measurements were spread overan extended period, three different versions of this software wereused: version 1.3.2.0, version 1.6 and version 1.7.4. The fluorescence amplitude threshold, distinguishing the positive from thenegative droplets was set manually by the analyst as the midpointbetween the average fluorescence amplitude of the positive andnegative droplet cluster. The same threshold was applied to all thewells of one PCR plate. Measurement results of single PCR wellswere excluded on the basis of technical reasons in case that (i) thetotal number of accepted droplets was 10,000, (ii) the average fluorescence amplitudes of positive or negative droplets were clearlydifferent from those of the other wells on the plate, or (iii) 5 % of theaccepted droplets had a fluorescence amplitude significantly belowthe average amplitude of the negative droplet cluster (i.e. average 4 STD). The average number of accepted droplets of the validmeasurement results was around 17,000.The numbers of positive and accepted droplets were transferredto an in-house developed spread sheet to calculate the copy numberconcentration in the sample (csample ) using Eq. (1) with a dropletvolume set at 0.834 nL [22].csample Dfsample DfPCR 31 1A Vd log 1 AP 1log 1 (1)AWith Dfsample : dilution factor of the DNA sample before addingto the PCR mix;DfPCR : dilution factor of the DNA solution in the PCR mix;A: number of analysed droplets;P: number of positive droplets;Vd : droplet volume.Throughout this manuscript, the term sample copy numberconcentration(csample ) is used to describe the copy number concentration of the undiluted sample, while the term PCR copy numberconcentration (cPCR ) is used to refer to the copy number concentration in the PCR mix.The dMIQE checklist [18] of these ddPCR experiments can befound in the Supplementary data Table 3.3. Results3.1. SelectivityThe primers and probes of the BCR-ABL ddPCR method are alsoused in a standardised qPCR method developed during a largeinter-laboratory study and the absence of nonspecific amplificationartefacts in qPCR has been shown [20,21]. The selectivity of the BCRABL ddPCR method was experimentally assessed by performing 4replicate measurements of a matrix blank consisting of 1 T1 E0.01buffer with the nucleic acid background of the ERM-AD623 samples (i.e. transfer RNA from E. coli) and 4 replicates of a positivecontrol consisting of an undiluted sample of ERM-AD623a at a PCRcopy number concentration of 54000 cp/ L. Results showed a cleardifference in fluorescence amplitude between the negative dropletcluster (average 1764 and STD 135) and the positive droplet cluster(average 5418 and STD 212). With the threshold placed at the mid-The working range of the BCR-ABL ddPCR method was investigated by measuring one sample of each of the five lowestERM-AD623 concentration levels at different PCR copy numberconcentrations: ERM-AD623b was measured at 5400 cp/ L, ERMAD623c at 2575 cp/ L, ERM-AD623d at 255 cp/ L, ERM-AD623e at26 cp/ L and ERM-AD623f at 2.5 cp/ L. For each concentration, 8replicate measurements were performed, and the replicates werespread over 4–5 cartridges and randomly positioned on the 96-wellPCR plate.None of the measurement results was rejected based on thetechnical reason exclusion criteria described in Section 2.3. Therelative STD of the replicate measurement results was 5% for thePCR copy number concentrations between 26 and 5400 cp/ L. Atthe lowest PCR copy number concentration of 2.5 cp/ L, the relative STD increased to 16.9 % suggesting that this concentrationmight be out of the working range. A precise determination of thelower end of the working range is discussed during the assessmentof the limit of quantification (LOQ) of the method in Section 3.6.The relation between the expected PCR copy number concentration (cPCR,exp ) and the measured PCR copy number concentration(cPCR,meas ) was linear (r2 0.9985, see Fig. 1) and the equation ofthe regression line was cPCR,meas 0.8677 cPCR,exp . This regressionline indicates that cPCR,meas is about 13 % lower than the cPCR,exp suggesting a bias between the certified copy number concentrations ofERM-AD623 and the copy number concentrations measured by theBCR-ABL ddPCR method. A more precise estimate of this bias basedon many measurement results was obtained during the assessmentof the method accuracy below.3.3. AccuracyThe five highest concentration levels of ERM-AD623 were measured with the BCR-ABL ddPCR method at PCR copy numberconcentrations of 250–450 cp/ L (for ERM-AD623a, b, c and d)and 25–35 cp/ L (for ERM-AD623e). The samples of ERM-AD623a,ERM-AD623b and ERM-AD623c were gravimetrically diluted inT1 E0.01 buffer to a nominal concentration between 1000 cp/ L and1800 cp/ L before adding to the PCR Mix. The experiments wereperformed in 3 runs and each ERM-AD623 concentration level wasmeasured with 12 replicates in runs 1 and 3, and 16 replicates inrun 2. The replicate measurements within one run were carriedout under repeatability conditions meaning: the same analyst, thesame pre-sample mix, cartridges from the same batch, the sameinstruments and randomly positioned on the same 96-well PCRplate. Between the runs, intermediate precision conditions wereapplicable, meaning: 3 different analysts, 2 different droplet generators, 2 different droplet readers, 2 different types of cartridges, 3different batches of reagents and 3 different versions of the QuantaSoft software. In total, 200 measurement results were obtained,and only 4 of them were rejected because of technical reasons.The nested design of this experiment allowed an estimation ofthe method repeatability and the run-to-run variation as prescribedby ISO 5725-3. [23] The results were grouped per run and analysedwith a one-way analysis of variance (ANOVA) test. For each ERMAD623 concentration level, the relative repeatability (srepeat,rel ) and
32L. Deprez et al. / Biomolecular Detection and Quantification 9 (2016) 29–39Fig. 1. Linearity of the BCR-ABL ddPCR method when measuring ERM-AD623 samples within a PCR copy number concentration range of 2.5 cp/ L to 5400 cp/ L.The data points represent the average result for eight replicate measurements, and the vertical error bars represent the associated STD. The horizontal error bars representthe standard uncertainty associated with the certified values of the ERM-AD623 samples.the relative run-to-run variation (srun,rel ) both expressed as STDwere calculated using Eqs. (2) and (3), respectively. srepeat,rel srun,rel MSwithinrun(2)c̄sample,meascannot be estimated with Eq. (3). In this case, we considered srun,relequal to zero as it is negligible compared to the srepeat,rel .Based on the srepeat,rel and srun,rel the relative standard uncertainty of the method precision (uprecision,rel ) associated with theaverage measured sample copy number concentration (c̄sample,meas )was calculated using Eq. (4). MSbetweenrun MSwithinrunn̄repli(3)c̄sample,measWith MSwithin run : the within run mean of squares calculated byone-way ANOVAMS between run : the between run mean of squares calculated byone-way ANOVAn̄repli : average number of replicates per runc̄sample,meas : average measured sample copy numberconcentration over all runs.It should be noted that srepeat,rel and srun,rel are estimates of thetrue STD and are subject to random fluctuations. It can, therefore,happen that MSbetweenrun is smaller than MSwithinrun and then srun,reluprecision,rel 2srepeat,reln̄repli nrun 2srun,rel(4)nrunWith nrun : number of runs, which is 3 in this caseThe calculated values for srepeat,rel , srun,rel , and uprecision,rel areshown in Table 2. The five values obtained for each parameter(one per ERM-AD623 concentration level) were combined into onepooled value by taking the root mean square (RMS, also calledquadratic mean) calculated as the square root of the average of thesquared values. The pooled relative repeatability (srepeat,pooled,rel )was 6.1 %, the pooled relative run-to-run variation (srun,pooled,rel )was 2.9 % and the pooled relative standard uncertainty related toprecision (uprecision,pooled,rel ) was 1.9 %.Table 2Results of the accuracy assessment of the BCR-ABL ddPCR method performed by measuring the five highest concentrations levels of ERM-AD623.CRMcsample,cert Ucert (cp/ L)c sample,meas (cp/ L)biasrel (%)srepeat,rel (%)srun,rel (%)uprecision,rel (%)AD623aAD623bAD623cAD623dAD623e(1.08 0.13) 10(1.08 0.11) 105(1.03 0.10) 104(1.02 0.09) 103(1.04 0.10) 1020.97 100.93 1050.94 1040.93 1030.97 102 10.2 13.8 9.0 8.5 sample,cert : certified sample copy number concentration,Ucert : expanded uncertainty of the certified copy number concentration, c sample,meas : average measured sample copynumber concentration, biasrel : relative bias between the certified value and the measured value, srepeat,rel : relative repeatability, srun,rel : relative run-to-run variation, uprecision,rel :relative standard uncertainty related to the precision, *: MSbetweenrun MSwithinrun .
L. Deprez et al. / Biomolecular Detection and Quantification 9 (2016) 29–39The trueness of the BCR-ABL ddPCR method was evaluated byestimating the relative bias (biasrel ) for each ERM-AD623 concentration level as the relative difference between the averagemeasured sample copy number concentration (c̄sample,meas ) and thecertified sample copy number concentration (csample,cert )(see Eq.(5)).biasrel c̄sample,meas csample,cert3.4. Measurement uncertaintytration level), was 9.6 %. To evaluate whether or not this biasrelis significant, the uncertainty associated with this bias estimatewas calculated taking into account the uncertainty associatedwith the average measured copy number concentrations (i.e.uprecision,pooled,rel ) and the relative standard uncertainty associatedwith the certified copy number concentration of each ERM-AD623concentration level (ucert,rel ). Both uncertainty contributions werecombined in the relative uncertainty of the bias estimate (ubias,rel )using Eq. (6).ei auprecision,pooled,rel 2 (ucert,rel,i )2ncert(6)With ncert : the number of certified reference materials used in theassessment of the bias.The ubias,rel was 5.4 %, and the relative expanded uncertainty ofthe bias estimate (Ubias,rel ) was calculated to be 10.9 % using Eq. (7).Ubias,rel 2 ubias,relAs the absolute value of the estimated biasrel is smaller than Ubias,relthis bias cannot be considered significant, but there is a strongindication that the BCR-ABL ddPCR method has the tendency tomeasure lower copy number concentrations than the chip-baseddPCR method used for the certification of the copy number concentration of the ERM-AD623 solutions.(5)csample,certThe average relative bias (biasrel ), calculated as the arithmeticmean of the five values for biasrel (one per ERM-AD623 concen-ubias,rel 33(7)Measurement uncertainty may arise from many sources and acomplete list of all potential sources is a good starting point for acomprehensive estimate of the overall measurement uncertainty[24]. Fig. 2 gives a schematic overview of all factors which maycontribute to the uncertainty of the measurement results obtainedwith the BCR-ABL ddPCR method as performed here.The results from the assessment for the method precisionprovided an estimate of the contribution of several uncertaintysources. The uncertainty contributions of the random effects,including sampling, random variation in the droplet volume,binominal distribution and the position in the thermocycler, wereincluded in the srepeat,pooled,rel while the srun,pooled,rel covers theuncertainty arising from the run-to-run effects such as the type ofcartridges, the reagent batches, the droplet reader/generator andthe analyst.The tendency to measure with the BCR-ABL ddPCR methodlower copy number concentrations than the certified copy number concentrations of the ERM-AD623 samples indicates that someof the remaining sources also have an important effect on themeasurement result and make a significant contribution to theoverall measurement uncertainty. These factors were thereforeinvestigated in greater detail. The estimation of the uncertaintyFig. 2. A schematic overview of all factors which may contribute to the uncertainty of the measurement results obtained with BCR-ABL ddPCR method as performed in thisvalidation study.Cprimers/probe : concentration primers and probe, Dfsample : dilution factor of sample before addition to PCR mix, DfPCR : dilution factor of sample in the PCR mix, Mdil : mass ofdiluent, Mdil sample : mass of diluent and sample, Mpremix : mass of pre sample mix, Mmix : mass of the PCR mix with sample, Vd : volume of the droplets
34L. Deprez et al. / Biomolecular Detection and Quantification 9 (2016) 29–39contribution of several individual factors was based on previousknowledge and uncertainty components 1 % were not consideredas significant. These negligible uncertainty sources are the accuracyof the weighing, the uncertainty associated with density determination of the pre-sample mix and the uncertainty related to thepurity and quality of the HPLC-purified primers and probes [25].The samples that are intended to be measured with the BCR-ABLddPCR method are highly purified solutions of linearized plasmidDNA in a T1 E0.01 buffer with a nucleic acid background. As thesesolutions are candidate certified reference materials, the intactness of the DNA molecules and their stability has already beeninvestigated. Due to the particular nature of the samples, the following sources of uncertainty were also considered to be negligible:presence of single-stranded DNA, presence of PCR inhibitors, thepresence of secondary DNA structures, which might disturb therandom distribution of the target sequence over the droplets, andthe accessibility and intactness of the target sequence.The droplet volume determines the absolute copy number concentration calculated with Eq. (1). We have used a droplet volumeof 0.834 nL as this volume was previously measured in our laboratory using the same equipment, the same type of supermix and thesame type of samples. The relative standard uncertainty associatedwith the measured droplet volume (uVd , rel) was 1.8 % [22].Two sources of measurement uncertainty (i.e. the assay and thethreshold setting) were investigated in a dedicated study to estimate their contribution to the overall measurement uncertainty.3.4.1. Uncertainty component related to the assayBy measuring the ERM-AD623 samples with another combination of primers and probe, it has been investigated if the assay itselfhas a significant contribution to the measurement uncertainty.Therefore, the five highest concentration levels of ERM-AD623were also measured with the ABL ddPCR method. The set-up of theexperiments was identical to the experiments performed to assessthe accuracy of the BCR-ABL ddPCR method, meaning 3 runs witheach 12–16 replicates per ERM-AD623 concentration level underrepeatability conditions within the runs and intermediate precisionconditions between the runs. Fourteen of the 200 measurementresults obtained with the ABL ddPCR method were rejected becauseof technical reasons. The results of the ABL and the BCR-ABL assaywere grouped per assay and per ERM-AD623 concentration level.For each concentration, the relative STD due to the assay (sassay )was calculated using one way-ANOVA and Eq. (8). sassay,rel MSbetweenassay MSwithinassayn̄meas,assayc̄sample,meas(8)With MSwithin assay : the mean of square of results obtained withone assayMS between assay : the mean square between results obtained withthe two assaysn̄meas,assay : the average number of measurements per assayc̄sample,meas : average measured sample copy numberconcentration from both assaysThe five values for sassay (one per ERM-AD623 concentration level) were pooled by calculating the RMS. The resultingsassay,pooled,rel was 1.0 % indicating that the uncertainty contributionof the assay can be considered as negligible.3.4.2. Uncertainty related to the threshold settingThe classification of the droplets into positive or negativedepends on the fluorescence amplitude of the threshold. For theexperiments performed here, the threshold was set at the midpointbetween the average fluorescence amplitude of the positive andnegative droplet cluster. However, other approaches can be used,and they may lead to different measurement results. The variabil-ity among the results obtained with different threshold settings iscaused by the presence of droplets with fluorescence amplitudeabove the upper boundary of the negative cluster and below thelower boundary of the positive cluster, the so-called rain droplets.We defined the boundaries of the negative and positive dropletcluster as the average amplitude 4 STD as this range would theoretically include all droplets of that cluster in case of a normaldistribution of the fluorescence data. It is unclear whether or not therain droplets, in reality, contain a copy of the target sequence as weobserved rain droplets in both the matrix blank and the highly positive control sample. An estimation of the maximum uncertaintycontribution related to the threshold setting can be obtained byanalysing the same measurement data with 3 completely differentapproaches to classifying the rain droplets: Low threshold placed at the upper boundary of the negativedro
the PCR plate was placed in a C1000 TouchTM Thermal Cycler (Bio-Rad, cat no. 185-1197) for PCR amplification. The PCR proto-col can be found in the Supplementary data Table 2. The droplet reading was done with the QX 100 Droplet reader (Bio-Rad, cat no. 186-3001) using ddPCRTM Droplet Reader
Quanti-Tray Demonstration Add Colilert to sample and shake to dissolve Pour mixture into a Quanti-Tray. 13 25 Quanti-Tray Demonstration cont. Seal and then incubate at 35 C for 24 hours Count positive wells and refer to MPN table 26 E.coli- Blue Fluorescence- Quanti-Tray under a 365nm UV Light. 14 27File Size: 1MBPage Count: 23Explore furtherFecal coliform and E. coli Analysis in wastewater and .ohiowea.orgAddressing Total Coliform Positive or E.coli Positive .www.epa.govMethod 1604: Total Coliforms and Escherichia coli in Water .www.epa.gov5.11 Fecal Bacteria Monitoring & Assessment US EPAarchive.epa.govMost probable number (MPN) method for counting coliform .www.onlinebiologynotes.comRecommended to you based on what's popular Feedback
Quanti-Tray and Quanti-Tray/2000 Enzyme Substrate Methods Enzyme based methodologies detect both total coliforms and E. coli simultaneously. Easy, rapid, accurate Widely accepted as standard for microbiological analysis of water. Enzymes for Quanti-Tray method: Colilert, Colisure, Colilert-18File Size: 495KBPage Count: 10
Quanti-Tray* Enumeration Procedure 1. Add contents of one pack to a 100 mL water sample in a sterile vessel. 2. Cap vessel and shake until dissolved. 3. Pour sample/reagent mixture into a Quanti-Tray* or Quanti-Tray*/2000 and seal in an IDEXX Quanti-Tray* Sealer. 4. Place the sealed tray in a 35 0.5 C incubator for 24 hours. 5.
The IDEXX Quanti-Tray* Sealer PLUS is a motor-driven, heated-roller instrument designed to seal water samples inside IDEXX Quanti- Trays. When used with Quanti-Trays and any IDEXX reagent (such as Colilert*, Colilert*-18, Colisure*, Enterolert*, Pseudalert*, or HPC for
Tray IDEXX, y compris Quanti-Tray*/2000. Ce Sealer, utilisé avec des Quanti-Tray et tout réactif de IDEXX, tel que Colilert*, Colilert-18*, Colisure*, Enterolert*, Pseudalert* et HPC pour Quanti-Tray*, automatise l’introduction de l’échantillon pour l’énumération bactérienne.
1 Lab meeting and introduction to qualitative analysis 2 Anion analysis (demonstration) 3 Anion analysis 4 5. group cation anion analysis 5 4. group cation (demonstration) 6 4. group cation anion analysis 7 3. group cation (demonstration) 8 3. group cation anion analysis 9 Mid-term exam 10 2. group cation (demonstration)
Application of Statistical Image Analysis in Quanti cation of Neutron Radiography Images of Drying I.M. Fijaª-Kirejczyk a;, J.J. Milczarek a, J. oª¡dek-Noakw a, F.C. de Beer b, M.J. Radebe b and G. Nothnagel b aNational Centre for Nuclear Re
the quanti cation of aromaticity of azines [ ], choice of a suitable reference model is shown to play very crucial role due to anisotropic e ect [ ], arising from heteroatoms. We had compared di erent nonaromatic models (shown below in Figure ) and the best nonaromatic model for reliable estimate of quanti cation of aromaticity is the one with par-
