• We are sorry, but NCBI web applications do not support your browser and may not function properly. More information
Logo of narLink to Publisher's site
Nucleic Acids Res. Jun 15, 2002; 30(12): e60.
PMCID: PMC117302

Integration of DNA ligation and rolling circle amplification for the homogeneous, end-point detection of single nucleotide polymorphisms

Abstract

Association studies using common sequence variants or single nucleotide polymorphisms (SNPs) may provide a powerful approach to dissect the genetic inheritance of common complex traits. Such studies necessitate the development of cost-effective, high throughput technologies for scoring SNPs. The method described in this paper for the co-detection of both alleles of a SNP in a single homogeneous reaction combines the specificity of a high fidelity DNA ligation step with the power of rolling circle amplification. The incorporation of Amplifluor™ energy transfer primers enables signal detection in a homogeneous format, making this approach highly amenable to automation. The adaptation of the genotyping method for high throughput screening using conventional liquid handling systems is described.

INTRODUCTION

Current estimates suggest that single nucleotide polymorphisms (SNPs) occur approximately once in every 250–2000 bases and represent the most common form of genetic variation (1). A key aim of pharmacogenomic research is the association of sequence variation with heritable phenotypes. It is anticipated that the rapid identification of SNPs will accelerate the identification of disease genes by allowing researchers to search for genetic polymorphisms associated with disease-related phenotypes (25). It is hoped that genome-wide SNP scoring will permit better correlation of therapies with clinical outcomes and drug effects.

The prospect of using high-density maps of SNPs to study the genetic basis of disease and drug response has created the need for methods of SNP typing that are scaleable for high throughput, are accurate and are easy to automate while remaining economical (6).

There are a variety of drawbacks associated with current PCR genotyping methods (7). Gel-based assays such as restriction fragment length polymorphism, oligonucleotide ligation assay and mini-sequencing are all based upon fragment separation, which is labour intensive, requires skilled staff and is difficult to apply to high throughput genotyping studies. Non-gel-based high throughput genotyping technologies include Taqman, pyrosequencing, microarray and mass spectrometry based methods. The pitfalls of these methods are that they either require allele-specific fluorescent probes that need extensive optimisation of reaction conditions, purification of the PCR sample is often necessary which limits throughput or an inflexible format is employed, which, in the case of microarrays, does not allow for the detection of both alleles in the same reaction. This has the potential to affect accuracy.

The integration of high fidelity ligation of allele-specific circular probes with subsequent detection by rolling circle amplification (RCA) as described in this paper provides a high throughput, isothermal method of target DNA amplification and detection. Using dual fluorescent labels, it enables the co-detection of both alleles of a SNP in a single homogeneous reaction.

Ligase-mediated detection of SNPs, followed by amplification of the signal by RCA, has been described either in the format of a solid phase/in situ assay (811) or as a fluorescent, biallelic assay that involves post-amplification, purification and labelling steps (12). The technique described in this paper enables the homogeneous, end-point detection of both alleles of a SNP in the same reaction, which improves genotyping accuracy by eliminating miscalls of heterozygotes due to reaction failures. The method requires two allele-specific circular probes to provide highly accurate allele discrimination. Generation of a positive signal depends on two hybridization events per probe combined with a ligation event, a process that confers a high degree of specificity to the assay. Generic amplification primers are used that do not require redesigning for each SNP, thereby reducing the cost per assay. A fluorescence resonance energy transfer based readout is employed which does not necessitate post-assay purification steps. All reactions are isothermal, making the assay amenable to automation and miniaturization within a microtitre plate format. This method therefore offers advantages in terms of throughput, cost and accuracy compared with current PCR genotyping methods.

The real-time identification of SNPs from genomic starting material has previously been described by Faruqi et al. (13). Here, we describe the adaptation of this method to overcome the problem of non-specific background signal that is sometimes related to primer artifacts or to the amount of probe used in the assay. Simple plate reader end-point detection and automation of the assay have been demonstrated from probe design to data analysis to provide a truly high throughput method for genotyping SNPs.

MATERIALS AND METHODS

Probe design

Adapting the procedure described by Faruqi et al. (13), for each SNP, two linear, allele-specific probes were designed to discriminate between the two alleles (Table (Table1).1). Probe design rules were incorporated into an algorithm which enabled automated design of allele-specific probes following input of the sequence content surrounding the SNP of interest.

Table 1.
Sequences of open circle probe backbones and amplification primers

Each probe consisted of an oligonucleotide that was 80–90 bases in length. At the 5′-end, the probe was phosphorylated and bore a stretch of at least 20 nt that hybridized to the complementary sequence on the template, immediately 5′ of the SNP. The 3′ end of the probe was designed to have at least 12 nt complementary to the region immediately 3′ of the SNP. The locus-specific regions of both open circle probes were identical with the exception of the 3′-terminal base, which was varied to complement each allele. Sandwiched between the locus-specific probe sequences was a backbone sequence of ~50 nt that encoded the binding sites for two RCA amplification primers, one of which was common to the two allele-specific probes. The other was probe-specific to enable discrimination (Table (Table11).

The first stage of the assay involved hybridisation of the probes to the target DNA. A stable hybrid was only formed if the 3′ base of the probe was perfectly matched to the polymorphic nucleotide present in the target. Thermostable DNA ligase (Ampligase™; Epicentre Technologies, Madison, WI) preferentially ligated perfectly matched hybrids (Fig. (Fig.1).1). Detection of covalently circularised oligonucleotide probes was performed by RCA using one common and one non-cross-reactive primer for each probe. The first amplification primer hybridised to the probe and in the presence of a strand-displacing DNA polymerase, the primer was extended, until it eventually displaced itself at its 5′ end once one complete revolution of the circularised probe was made. Continued polymerisation and displacement resulted in the generation of a single-stranded concatamer known as the first strand RCA product.

Figure 1
Schematic of RCA for SNP identification. The allele-specific probes hybridise to target DNA so that the 3′-terminal nucleotide is at the SNP position. Ligation and therefore circularisation of the probes only occurs if the 3′-terminal ...

A second primer of the same sense as the original probe was able to recognise and bind to its cognate target site in each tandem repeat of the RCA first strand product and initiate DNA synthesis at each site. As these multiple binding events elongated, strand displacement caused these nascent strands to be displaced and a multitude of new recognition sites for the first primer were exposed for further priming and extension.

Strand displacement eventually resulted in the release of double-stranded DNA fragments from the parent molecule. These displaced DNA molecules accumulated as a nested population of fragments, the size of each successive fragment increasing by the length of the original open circle probe. This geometric amplification, involving ever increasing and self-propagating strand displacement and DNA fragment accumulation, is called hyperbranching (8) (Fig. (Fig.11).

The second amplification primer was designed as an Amplifluor primer labelled with a 5′-hairpin structure bearing a fluorophore and quencher (14). The fluorophore and quencher were in sufficient proximity in the native primer for fluorescence quenching to occur. During incorporation into double-stranded RCA products, the polymerase read through the hairpin and opened out the stem–loop. In this linearised, double-stranded conformation the fluorophore was separated and no longer quenched so that fluorescence could be detected (Fig. (Fig.1).1). The sense primers to each probe backbone were labelled with two different fluorophores, enabling discrete detection of both alleles of a SNP in a single reaction.

Assay automation

The ligation-dependent RCA assay was adapted to an automated protocol which addressed probe design, liquid handling and data analysis. Probe design rules were incorporated into a software algorithm that was developed in-house. This permitted automated design of allele-specific probes following input of the sequence content surrounding the SNP of interest. A combination of off-the-shelf and custom liquid handling and thermal incubation solutions were devised for both 96- and 384-well microtitre plates based around the Tecan Genesis RSP200 liquid handling station (Tecan Austria GmbH, Grödig, Austria). These were linked by robotic arms (Genesis robot manipulator arm, RoMa) which enabled throughputs of 10 000 SNPs/day. Data analysis software for SNP scoring was developed in-house and was based upon a modified c-means clustering algorithm (15), which provided quality-related indices that included a degree of certainty to which cluster partitioning was correct in addition to individual doubt measures for the assignment of each data point to a cluster. These values provided a quality measure for individual SNP calls (F.Ghouze, G.Scozzafava, R.Oreo, B.Hughes, P.Roe, C.Wheeler, R.Howe, S.Morris and J.Comley, manuscript submitted).

Template DNA

Samples previously genotyped for three SNPs, human apolipoprotein C-II (APOC2, accession no. M10612, G/T SNP at nucleotide 522), human β-1 integrin (ITGB1, accession no. X68969, O/G SNP nucleotide 590) and human pyruvate dehydrogenase kinase isoenzyme 2 (PDK2, accession no. L42451, G/C SNP at nucleotide 1357), were amplified by PCR. Each PCR was performed in a 100 µl reaction volume, containing 100 ng genomic DNA, 100 pmol each primer, 200 µM dNTP and 1.0–2 mM MgCl2. Thermal cycling was performed in a PTC-225 DNA machine (MJ Research Inc., Cambridge, MA) at 95°C for 15 min, followed by 30 cycles of 94°C for 30 s, annealing temperature for 30 s and 72°C for 60 s, then by 72°C for 5 min. The annealing temperatures were 45°C for APOC2, 61°C for ITGB1 and 62°C for PDK2.

Ligation

Each ligation reaction was carried out in a 384-well ThermoFast mark II plate (ABgene, Epsom, UK), in a total volume of 10 µl, containing 107–109 copies of PCR template, 50 nM each open circle probe (Table (Table1),1), 20 mM Tris–HCl pH 8.3, 25 mM KCl, 0.01% Triton X-100, 3 mM magnesium chloride, 500 µM NAD+ and 0.625 U Ampligase thermostable DNA ligase. The plate was sealed with a Microseal ‘A’ film (MJ Research, Waltham, MA) and incubated for 3 min at 95°C, followed by 65°C for a further 20 min, then rapidly ramped to 4°C prior to the addition of reagents required for the amplification and detection of ligated probes.

RCA reactions

The volume of each ligation reaction was increased to 20 µl by the addition of 0.6 µM antisense primer, 0.1 µM each sense primer A and B, 20 mM Tris–acetate pH 8.3, 10 mM ammonium sulphate, 200 µM dNTPs and 5.36 U Bst DNA polymerase large fragment (New England Biolabs, Beverly, MA). The reactions were incubated for 60 min at 65°C before ramping quickly to 4°C.

Detection and analysis

The plate containing the completed reactions was placed into a plate support (ABgene), then scanned at 530/590 nm (Cy3 excitation/emission) and 485/535 nm (FAM excitation/emission) using a Tecan Ultra plate reader. The resulting Cy3 and FAM signals from each well were plotted against one another on a scatter plot. The data were then analysed with software that was developed using a modified fuzzy c-means clustering algorithm (15). This associates each data point to a specific genotype and generates confidence parameters that could be utilised to assign data quality measures. The accuracy of SNP scoring using this RCA method was determined by comparing the resulting genotypes with known genotypes derived from DNA sequencing data.

The principle of this RCA method has been demonstrated using three SNPs in the APOC2, ITGB1 and PDK2 genes (Table (Table22).

Table 2.
Sequences of the open circle probes designed to interrogate the APOC2, ITGB1 and PDK2 SNPs

Testing primer and backbone sequences

To test for the ability of the amplification primers to form artifacts, ligation and RCA reactions were performed as described above, but in the absence of target and open circle probe (i.e. the only DNA present was primer DNA). Positive control reactions were performed in the presence of target and open circle probe for the APOC2 SNP. The primers were unlabelled, so amplicons were detected by the addition of 12× Gelstar nucleic acid stain (BioWhittaker Molecular Applications, Rockland, ME) to the RCA reaction. The incubation and analysis was performed in a Smart Cycler (Cepheid, CA). The resulting real-time data were used to determine the time between the appearance of positive signal and the appearance of background.

To compare the amplification efficiencies of wild-type and mutant open circle probes for each SNP, each probe was ligated independently of the other, then amplified by unlabelled primers in the presence of 12× Gelstar nucleic acid stain. The incubation and analysis were performed in a Smart Cycler. The resulting real-time data were used to determine the time taken to reach half of the maximum fluorescence (CT) for each probe. The difference between the CT value for each probe pair (ΔCT) was calculated as an average of three experiments (n = 3).

RESULTS

Robustness of open circle probe design

The ability of open circle probes to differentiate between the two alleles of a SNP was tested in the presence of probes complementary to both alleles. Synthetic oligonucleotide targets were used in ligation and RCA reactions to mimic both homozygous and heterozygous genotypes. Successfully designed probes were able to differentiate the alleles correctly. Probes have been designed for 99 SNPs, ranging between 20 and 84% GC content of the arm sequences. The results show that at the first design attempt, 95% of probes demonstrated successful allelic discrimination (defined by the generation of distinct clusters for each genotype tested; Fig. Fig.22).

Figure 2
Accuracy of open circle probe design. For every SNP, both allele-specific probes are ligated and amplified in the presence of synthetic oligonucleotide targets of known sequence. Successfully designed probes were able to accurately discriminate between ...

Testing primer and backbone sequences

Primers were incubated in the presence or absence of heterozygous APOC2 PCR and open circle probe to determine at what point non-specific primer artifacts began to appear. The RCA reactions were monitored in real time and the resulting data were used to determine the size of window between the generation of genuine signal and primer artifact. On average a positive signal was detected at 24 min (n = 3) and signal from non-specific background artifacts was not detected until 99 min into the reaction (n = 3). This demonstrates an average window of 75 min between detection of positive signal and the generation of non-specific background.

The open circle probe backbones and their associated amplification primers were designed to promote equal amplification efficiencies and therefore to minimise differential amplification between the two alleles of a SNP that might result in miscalls. The kinetics of probe amplification for the three SNPs were assessed by ligating and amplifying each probe separately in the presence of heterozygous target. DNA synthesis, monitored by fluorescence from an intercalating dye, was measured in real time at 3 min intervals over a period of 90 min. Plots of the fluorescence against time were used to determine the time taken to reach maximum fluorescence (CT) and therefore half of the maximum fluorescence (CT) of each probe. The results (Table (Table3)3) show that the CT for pairs of open circle probes ranged between 22.5 and 33 min. The difference between the CT value for a probe pair (ΔCT) ranged between 0.8 and 5 min dependent upon the SNP in question. This suggests that the probe kinetics for each SNP are well matched. This compares to the earliest appearance of non-specific background signal at 84 min in a negative control sample (the standard assay is only incubated for 60 min).

Table 3.
Comparison of time taken to reach half of the maximum fluorescence signal (CT) of probes 1 and 2 for the APOC2, ITGB1 and PDK2 SNPs

Demonstration of accurate SNP scoring

The chosen three SNPs represent a typical SNP (APOC2), an insertion/deletion polymorphism (ITGB1) and a SNP surrounded by sequence rich in GC residues (PDK2) (Table (Table22).

Open circle probe pairs were designed for each and more than 60 DNA samples of known genotype were tested. For each SNP, 24 negative control reactions were included (eight without probe, eight with no ligase and eight lacking target DNA). Ligation and RCA reactions were performed and the resulting FAM and Cy3 fluorescent signals from each well plotted on a cluster graph (Fig. (Fig.3).3). Samples were analysed with software that uses a modified fuzzy c-means clustering algorithm, where clustering is based on an angular representation of the data points. Each cluster therefore represents a genotype. Points close to the origin are those that emit negligible fluorescent signal, e.g. negative controls or failed reactions. The software therefore rejected these points. The assays were performed blind and the accuracy of calling was determined by comparing the predicted genotypes with known genotypes.

Figure 3
Scatter plots of the RCA end-point FAM and Cy3 fluorescence values for the APOC2 (top), ITGB1 (middle) and PDK2 (bottom) SNPs. The position of each data point is determined by the relative amplification of each probe within that reaction, which only ...

In all cases, for each SNP the negative controls produced a negative call when analysed by the software. For SNPs APOC2 and PDK2, 68 and 64 DNA samples were tested respectively. In both cases 100% of reactions were called correctly when the predicted genotypes were revealed. Detection of the SNP in INTB1 templates was tested in 60 samples. Upon analysis, four of these 60 samples were given a ‘no-call’ status because very little signal was generated by these samples following RCA amplification. The remaining 56 samples were called correctly. No miscalls or false positives were generated. Data from these experiments demonstrate the 100% accuracy of this method. Of the 192 samples tested, four failed and no miscalls were given.

CONCLUSIONS

We report a new fluorescent end-point assay for SNP scoring which simultaneously detects both alleles in a single reaction. The method utilises the highly specific properties of a thermostable DNA ligase to circularise oligonucleotide probes and thus discriminate a match or mismatch between the 3′ base of the probe and the target SNP. Ligated probes are subsequently amplified and detected by RCA with primers that emit a fluorescent signal only when incorporated into the amplified product. Using two fluorescently labelled primers, each directed to one probe backbone, it is possible to differentiate between the amplified products of both allele-specific probes in the same reaction, which improves genotyping accuracy by eliminating miscalls of heterozygotes due to reaction failures. A homogeneous fluorescent end-point also eliminates the need for purification or separation steps enabling a higher level of throughput. The entire technique is therefore highly amenable to automation because few pipetting steps are required and the incubations, which take no longer than 90 min in total, are isothermal. This method has been successfully automated to cover a range of throughput capacities (F.Ghouze, G.Scozzafava, R.Oreo, B.Hughes, P.Roe, C.Wheeler, R.Howe, S.Morris and J.Comley, manuscript submitted).

Good allelic discrimination is highly dependent upon the design of the probe arms. It is important to choose a melting temperature (Tm) for the probe 5′ arm that is close to the reaction temperature, i.e. 65°C. This ensures that the 5′ arm is able to form a stable association with the sequence adjacent to the SNP of interest. Discrimination is aided by designing the 3′ arm with a Tm of ~40°C. This promotes a continuous cycle of hybridisation and melting that functions to improve specificity and hence discrimination between match and mismatch of the 3′ base of the probe and the SNP site. Ligation of adjacent sequences with a mismatch at the 3′ end is markedly poor. The efficiency of the thermophilic ligase Thermus thermophilus DNA ligase for ligating matched sequences is reported to be 5.9 × 103 greater than that for 3′ end mismatched sequence ligation (16). In a homozygous sample, ligation of one of the allele-specific probes therefore predominates over the other. In a heterozygous sample, both open circle probes are ligated with similar efficiencies. The probes were also designed to avoid positioning a guanine (G) at the 3′ end, because mismatches containing a guanine are discriminated less efficiently by the ligase used than other bases. In such cases, probes were designed to the opposite strand of the template DNA, e.g. a G/T mismatch would become a C/A mismatch. Adhering to these design rules has resulted in a 95% success rate at the first attempt of probe design.

Careful design of the open circle probes and primer systems is critical to the success of this technique and is necessary to achieve matched amplification kinetics that display minimal levels of target-independent, non-specific amplification, a problem reported by others in the field (13). OLIGO™ 6 (Molecular Biology Insights Inc., MN) and Lasergene™ primer design software (DNAStar Inc., WI) were used to generate potential primer sequences from multiple 1 Mb blocks of randomly generated DNA sequence. Selection criteria included the requirement for a GC clamp at the 5′ end of each primer candidate and a 3′-terminal adenosine. The primers were designed to have a Tm close to 65°C (the reaction temperature) and minimal self-dimerisation or annealing to other potential primers expected to be present in the reaction. Successful primer sequences were entered into a BLAST (National Center for Biotechnology Information, MD) search to eliminate those with significant homologies with known human or eukaryotic sequences. The surviving oligonucleotide sequences were used to construct the probe backbones. A selection of backbones and primers were screened for combinations that gave optimal RCA efficiency in the absence of non-specific background amplification.

Strategies to rescue failed probes have yet to be investigated fully, but could involve designing probes to the opposite strand of the target sequence or the use of universal or degenerate bases when more than one SNP is located within the region of interest. A small percentage of SNPs, however, may well be refractory to analysis (due, for example, to a high degree of secondary structure in the arm sequences or repeat sequences). An instance of the latter is a SNP in human chromosome 19 (Human Genome draft contig NT_011157.4). This SNP is located within a block of sequence that is perfectly duplicated 91.5 kb downstream (nucleotide positions 478335 and 569893). It would therefore not be possible to score the SNP by any current method. In such cases an alternative SNP site should be sought.

The accuracy of this SNP scoring method has been illustrated using three SNPs, tested in a total of 264 reactions (inclusive of unknown samples and controls). Accuracy is 100% for all SNPs tested, including the PDK2 SNP surrounded by GC-rich sequence and the insertion/deletion polymorphism in ITGB1. The genotyping data for each SNP is grouped into distinct clusters when plotted on scatter graphs. The clusters were analysed with software that applies a modified fuzzy c-means clustering algorithm to the data, enabling accurate SNP typing by associating each cluster to a specific genotype. This algorithm was chosen because it generates confidence parameters that can be utilised to assign data quality measures. The algorithm was modified to deal with data that tends not to group as point clusters but instead as linear clusters radiating from the point of origin. This is caused partly by variations in the number of target copies between individual samples, so that samples containing fewer copies of target will be amplified to a lesser extent than samples containing higher numbers of target copies. In addition, it is not possible to discount the presence of a low level of intrinsic variability. The results show that no samples were miscalled. Four samples gave no calls, possibly due to degradation of target. The ability of this method to multiplex the detection of both alleles of a SNP in the same reaction leads to greater robustness. It is possible to produce highly accurate data as shown here, where accuracies of 100% have been achieved for the three SNPs tested.

REFERENCES

1. Sachidanandam R., Weissman,D., Schmidt,S.C., Kakol,J.M., Stein,L.D., Marth,G., Sherry,S., Mullikin,J.C., Mortimore,B.J., Willey,D.L., Hunt,S.E., Cole,C.G., Coggill,P.C., Rice,C.M., Ning,Z., Rogers,J., Bentley,D.R., Kwok,P.Y., Mardis,E.R., Yeh,R.T., Schultz,B., Cook,L., Davenport,R., Dante,M., Fulton,L., Hillier,L., Waterston,R.H., McPherson,J.D., Gilman,B., Schaffner,S., Van Etten,W.J., Reich,D., Higgins,J., Daly,M.J., Blumenstiel,B., Baldwin,J., Stange-Thomann,N., Zody,M.C., Linton,L., Lander,E.S. and Atshuler,D. The International SNP Map Working Group (2001) A map of human genome sequence variation containing 1.42 million single nucleotide polymorphisms. Nature, 409, 928–933. [PubMed]
2. Collins F.S., Brooks,L.D. and Chakravarti,A. (1998) A DNA polymorphism discovery resource for research on human genetic variation. Genome Res., 8, 1229–1231. [PubMed]
3. Kwok P.Y., Deng,Q., Zakeri,H., Taylor,S.L. and Nickerson,D.A. (1996) Increasing the information content of STS-based genome maps: identifying polymorphisms in mapped STSs. Genomics, 31, 123–126. [PubMed]
4. Brookes A.J. (1999) The essence of SNPs. Gene, 234, 177–186. [PubMed]
5. Riley J.H., Allan,C.J., Lai,E. and Roses,A. (2000) The use of single nucleotide polymorphisms in the isolation of disease genes. Pharmacogenomics, 1, 39–47. [PubMed]
6. McCarthy J.J. and Hilfiker,R. (2000) The use of single-nucleotide polymorphism maps in pharmacogenomics. Nat. Biotechnol., 18, 505–508. [PubMed]
7. Shi M.M. (2001) Enabling large-scale pharmacogenetic studies by high-throughput mutation detection and genotyping technologies. Clin. Chem., 47, 164–172. [PubMed]
8. Lizardi P.M., Huang,X., Zhu,Z., Bray-Ward,P., Thomas,D.C. and Ward,D.C. (1998) Mutation detection and single-molecule counting using isothermal rolling-circle amplification. Nature Genet., 19, 225–232. [PubMed]
9. Nilsson M., Krejci,K., Koch,J., Kwiatkowski,M., Gustavsson,P. and Landegren,U. (1997) Padlock probes reveal single-nucleotide differences, parent of origin and in situ distribution of centromeric sequences in human chromosomes 13 and 21. Nature Genet., 16, 252–255. [PubMed]
10. Nilsson M., Malmgren,H., Samiotaki,M., Kwiatkowski,M., Chowdhary,B.P. and Landegren,U. (1994) Padlock probes: circularizing oligonucleotides for localized DNA detection. Science, 265, 2085–2088. [PubMed]
11. Nallur G., Luo,C., Fang,L., Cooley,S., Dave,V., Lambert,J., Kukanskis,K., Kingsmore,S., Lasken,R. and Schweitzer,B. (2001) Signal amplification by rolling circle amplification on DNA microarrays. Nucleic Acids Res., 29, e118. [PMC free article] [PubMed]
12. Qi X., Bakht,S., Devos,K.M., Gale,M.D. and Osbourn,A. (2001) L-RCA (ligation-rolling circle amplification): a general method for genotyping of single nucleotide polymorphisms (SNPs). Nucleic Acids Res., 29, e116. [PMC free article] [PubMed]
13. Faruqi A.F., Hosono,S., Driscoll,M.D., Dean,F.B., Alsmadi,O., Bandaru,R., Kumar,G., Grimwade,B., Zong,Q., Sun,Z., Du,Y., Kingsmore,S., Knott,T. and Lasken,R.S. (2001) High-throughput genotyping of single nucleotide polymorphisms with rolling circle amplification. BMCGenomics, 2, 4. [PMC free article] [PubMed]
14. Thomas D.C., Nardone,G.A. and Randall,S.K. (1999) Amplification of padlock probes for DNA diagnostics by cascade rolling circle amplification or the polymerase chain reaction. Arch. Pathol. Lab. Med., 123, 1170–1176. [PubMed]
15. Höppner F., Klawonn,F., Kruse,R. and Runkler,T. (1999) Fuzzy Cluster Analysis. John Wiley & Sons, Chichester, UK.
16. Luo J., Bergstrom,D. and Barany,F. (1996) Improving the fidelity of Thermus thermophilus DNA ligase. Nucleic Acids Res., 24, 3071–3078. [PMC free article] [PubMed]

Articles from Nucleic Acids Research are provided here courtesy of Oxford University Press
PubReader format: click here to try

Formats:

Related citations in PubMed

See reviews...See all...

Cited by other articles in PMC

See all...

Links

Recent Activity

Your browsing activity is empty.

Activity recording is turned off.

Turn recording back on

See more...