Limits...
Analysis of copy number variation using quantitative interspecies competitive PCR.

Williams NM, Williams H, Majounie E, Norton N, Glaser B, Morris HR, Owen MJ, O'Donovan MC - Nucleic Acids Res. (2008)

Bottom Line: Over recent years small submicroscopic DNA copy-number variants (CNVs) have been highlighted as an important source of variation in the human genome, human phenotypic diversity and disease susceptibility.Consequently, there is a pressing need for the development of methods that allow the efficient, accurate and cheap measurement of genomic copy number polymorphisms in clinical cohorts.In this study we establish proof of concept by accurately detecting previously characterized mutations at the PARK2 locus and then demonstrating the potential of quantitative interspecies competitive PCR (qicPCR) to accurately genotype CNVs in association studies by analysing chromosome 22q11 deletions in a sample of previously characterized patients and normal controls.

View Article: PubMed Central - PubMed

Affiliation: Department of Psychological Medicine, Wales School of Medicine, Cardiff University, Heath Park, Cardiff CF14 4XN, UK. williamsnm@cf.ac.uk

ABSTRACT
Over recent years small submicroscopic DNA copy-number variants (CNVs) have been highlighted as an important source of variation in the human genome, human phenotypic diversity and disease susceptibility. Consequently, there is a pressing need for the development of methods that allow the efficient, accurate and cheap measurement of genomic copy number polymorphisms in clinical cohorts. We have developed a simple competitive PCR based method to determine DNA copy number which uses the entire genome of a single chimpanzee as a competitor thus eliminating the requirement for competitive sequences to be synthesized for each assay. This results in the requirement for only a single reference sample for all assays and dramatically increases the potential for large numbers of loci to be analysed in multiplex. In this study we establish proof of concept by accurately detecting previously characterized mutations at the PARK2 locus and then demonstrating the potential of quantitative interspecies competitive PCR (qicPCR) to accurately genotype CNVs in association studies by analysing chromosome 22q11 deletions in a sample of previously characterized patients and normal controls.

Show MeSH

Related in: MedlinePlus

Estimated copy number at 22q11 in 753 samples. Replicate experiments of a single multiplex qicPCR panel to determine hsCN at 22q11 in twenty 22q11DS patients (blue) and 733 controls (red).
© Copyright Policy - creative-commons
Related In: Results  -  Collection

License
getmorefigures.php?uid=PMC2553599&req=5

Figure 6: Estimated copy number at 22q11 in 753 samples. Replicate experiments of a single multiplex qicPCR panel to determine hsCN at 22q11 in twenty 22q11DS patients (blue) and 733 controls (red).

Mentions: Given our experience that the performance of some assays can be less robust when scaled up, we analysed a larger series of 753 samples, composed of 733 non-deleted controls and 20 new 22q11DS samples independent of those in which the assay was optimized, dispersed among the control samples. For each sample the hsCN at the 22q11 locus was determined as an average of five locus-specific test probes and five reference probes all located within the same multiplex reaction. All samples were analysed in duplicate and similar results were obtained from both experimental replicates (Figure 6). All 22q11DS samples could be clearly distinguished from the 20 control samples (silhouette score = 0.81) (Figure 6). The mean hsCN estimate of the 22q11DS samples was 0.96 (0.46min−1.48max), mean CV = 0.20 (0.14min−0.23max), while that of the non-deleted controls was 2.07 (1.68min−2.82max), mean CV = 0.06 (0.05min−0.07max). High-throughput genotyping assays are expected to be sufficiently robust to analyse large numbers of samples (drop-out rate <5%) while maintaining very low genotype error rates (<0.1%). Blind genotyping of our samples using a simple Excel (Microsoft) spreadsheet, which adhered to a simple semi-automated protocol whereby samples were called as carrying either 1 or 2 copy numbers if the hsCN determined by qicPCR fell within ± 40% of the expected integer (1 copy = 0.6–1.4, 2 copies = 1.6–2.4). Estimates of hsCN that fell outside these ranges were classed as genotype failures. Analysis of each single pass experiment resulted in >97% of samples being genotyped with 100% accuracy (a total of 46/1506 samples failed to genotype). Moreover, determining the hsCN as the average of the two experimental replicates of each multiplex panel further improved the genotyping quality, with >99.4% of samples genotyped with 100% accuracy (a total of 5/753 samples analysed in duplicate failed to be genotyped), suggesting that further improvements in the quality of quantitative genotyping of CNVs by qicPCR are possible simply by performing one replicate experiment.Figure 6.


Analysis of copy number variation using quantitative interspecies competitive PCR.

Williams NM, Williams H, Majounie E, Norton N, Glaser B, Morris HR, Owen MJ, O'Donovan MC - Nucleic Acids Res. (2008)

Estimated copy number at 22q11 in 753 samples. Replicate experiments of a single multiplex qicPCR panel to determine hsCN at 22q11 in twenty 22q11DS patients (blue) and 733 controls (red).
© Copyright Policy - creative-commons
Related In: Results  -  Collection

License
Show All Figures
getmorefigures.php?uid=PMC2553599&req=5

Figure 6: Estimated copy number at 22q11 in 753 samples. Replicate experiments of a single multiplex qicPCR panel to determine hsCN at 22q11 in twenty 22q11DS patients (blue) and 733 controls (red).
Mentions: Given our experience that the performance of some assays can be less robust when scaled up, we analysed a larger series of 753 samples, composed of 733 non-deleted controls and 20 new 22q11DS samples independent of those in which the assay was optimized, dispersed among the control samples. For each sample the hsCN at the 22q11 locus was determined as an average of five locus-specific test probes and five reference probes all located within the same multiplex reaction. All samples were analysed in duplicate and similar results were obtained from both experimental replicates (Figure 6). All 22q11DS samples could be clearly distinguished from the 20 control samples (silhouette score = 0.81) (Figure 6). The mean hsCN estimate of the 22q11DS samples was 0.96 (0.46min−1.48max), mean CV = 0.20 (0.14min−0.23max), while that of the non-deleted controls was 2.07 (1.68min−2.82max), mean CV = 0.06 (0.05min−0.07max). High-throughput genotyping assays are expected to be sufficiently robust to analyse large numbers of samples (drop-out rate <5%) while maintaining very low genotype error rates (<0.1%). Blind genotyping of our samples using a simple Excel (Microsoft) spreadsheet, which adhered to a simple semi-automated protocol whereby samples were called as carrying either 1 or 2 copy numbers if the hsCN determined by qicPCR fell within ± 40% of the expected integer (1 copy = 0.6–1.4, 2 copies = 1.6–2.4). Estimates of hsCN that fell outside these ranges were classed as genotype failures. Analysis of each single pass experiment resulted in >97% of samples being genotyped with 100% accuracy (a total of 46/1506 samples failed to genotype). Moreover, determining the hsCN as the average of the two experimental replicates of each multiplex panel further improved the genotyping quality, with >99.4% of samples genotyped with 100% accuracy (a total of 5/753 samples analysed in duplicate failed to be genotyped), suggesting that further improvements in the quality of quantitative genotyping of CNVs by qicPCR are possible simply by performing one replicate experiment.Figure 6.

Bottom Line: Over recent years small submicroscopic DNA copy-number variants (CNVs) have been highlighted as an important source of variation in the human genome, human phenotypic diversity and disease susceptibility.Consequently, there is a pressing need for the development of methods that allow the efficient, accurate and cheap measurement of genomic copy number polymorphisms in clinical cohorts.In this study we establish proof of concept by accurately detecting previously characterized mutations at the PARK2 locus and then demonstrating the potential of quantitative interspecies competitive PCR (qicPCR) to accurately genotype CNVs in association studies by analysing chromosome 22q11 deletions in a sample of previously characterized patients and normal controls.

View Article: PubMed Central - PubMed

Affiliation: Department of Psychological Medicine, Wales School of Medicine, Cardiff University, Heath Park, Cardiff CF14 4XN, UK. williamsnm@cf.ac.uk

ABSTRACT
Over recent years small submicroscopic DNA copy-number variants (CNVs) have been highlighted as an important source of variation in the human genome, human phenotypic diversity and disease susceptibility. Consequently, there is a pressing need for the development of methods that allow the efficient, accurate and cheap measurement of genomic copy number polymorphisms in clinical cohorts. We have developed a simple competitive PCR based method to determine DNA copy number which uses the entire genome of a single chimpanzee as a competitor thus eliminating the requirement for competitive sequences to be synthesized for each assay. This results in the requirement for only a single reference sample for all assays and dramatically increases the potential for large numbers of loci to be analysed in multiplex. In this study we establish proof of concept by accurately detecting previously characterized mutations at the PARK2 locus and then demonstrating the potential of quantitative interspecies competitive PCR (qicPCR) to accurately genotype CNVs in association studies by analysing chromosome 22q11 deletions in a sample of previously characterized patients and normal controls.

Show MeSH
Related in: MedlinePlus