Improving the efficiency of genomic loci capture using oligonucleotide arrays for high throughput resequencing

Abstract

Background: The emergence of next-generation sequencing technology presents tremendous opportunities to accelerate the discovery of rare variants or mutations that underlie human genetic disorders. Although the complete sequencing of the affected individuals' genomes would be the most powerful approach to finding such variants, the cost of such efforts make it impractical for routine use in disease gene research. In cases where candidate genes or loci can be defined by linkage, association, or phenotypic studies, the practical sequencing target can be made much smaller than the whole genome, and it becomes critical to have capture methods that can be used to purify the desired portion of the genome for shotgun short-read sequencing without biasing allelic representation or coverage. One major approach is array-based capture which relies on the ability to create a custom in-situ synthesized oligonucleotide microarray for use as a collection of hybridization capture probes. This approach is being used by our group and others routinely and we are continuing to improve its performance.

Results: Here, we provide a complete protocol optimized for large aggregate sequence intervals and demonstrate its utility with the capture of all predicted amino acid coding sequence from 3,038 human genes using 241,700 60-mer oligonucleotides. Further, we demonstrate two techniques by which the efficiency of the capture can be increased: by introducing a step to block cross hybridization mediated by common adapter sequences used in sequencing library construction, and by repeating the hybridization capture step. These improvements can boost the targeting efficiency to the point where over 85% of the mapped sequence reads fall within 100 bases of the targeted regions.

Conclusions: The complete protocol introduced in this paper enables researchers to perform practical capture experiments, and includes two novel methods for increasing the targeting efficiency. Coupled with the new massively parallel sequencing technologies, this provides a powerful approach to identifying disease-causing genetic variants that can be localized within the genome by traditional methods.

Figure 1

Mapping of sequences relative to probe position in the genome. a) Sequence coverage distribution averaged across all targeted regions captured by basal capture protocol and b) sequence coverage distribution averaged across all targeted regions captured by double hybridization (modified) protocol show that the sequence reads are tightly limited around the targeted regions. Here, a targeted region is not necessarily a targeted exon but a probeset composed of multiple probes that are < 200 bp apart to each other. The y axis plots the relative abundance and the x axis is the base position relative to the probes positions.

Figure 2

Copy number fold differences between the normal and tumor tissues per chromosome using single hybridization capture protocol with blockers. The cancer specimen used in these experiments was known to have a chromosome 7 copy number gain and a chromosome 10 deletion. The normalized counts per chromosome are plotted for all chromosomes and are markedly different for the two chromosomes at altered copy numbers.

Figure 3

EGFR DNA amplification event is preserved in sequence data. A 200 Kb sized moving average of the interval flanking a) known EGFR amplification event are plotted in genomic position and b) for reference another genomic interval around the FOXP2 gene also on chromosome 7 is shown demonstrating the more typical coverage. The EGFR region is amplified 25× in average compared to the region outside of EGFR.

Figure 4

Percentage of targeted bases sequenced at various minimum coverage for different mean coverages. X-axis represents the coverage per base level and the corresponding y-axis represents the percentage of targeted bases that were covered at greater or equal with certain coverage. Table legends describe the detail of each line shown.