Abstract
Presentations at Select Biosciences’s first ‘Genomics Automation Congress’ (Boston, MA, USA) in 2010 focused on next-generation sequencing and the platforms and methodology around them. The meeting provided an overview of sequencing technologies, both new and emerging. Speakers shared their recent work on applying sequencing to profile cells for various levels of biomolecular complexity, including DNA sequences, DNA copy, DNA methylation, mRNA and microRNA. With sequencing time and costs continuing to drop dramatically, a virtual explosion of very large sequencing datasets is at hand, which will probably present challenges and opportunities for high-level data analysis and interpretation, as well as for information technology infrastructure.
The advent of next-generation sequencing, enabling much faster and cheaper sequencing of human genomes, is projected to revolutionize both biomedical research and clinical practice. In light of this, Select Biosciences (Sudbury, UK) held its first ‘Genomics Automation Congress’ in Boston (MA, USA) on 6–7 May 2010. The purpose of this conference was to look at some of the current and future technologies advancing the field of genomics, and this year the focus was entirely on next-generation sequencing and the platforms and methodology around it; 14 talks were given by leading experts in the field. The conference was co-located with both the ‘RNAi & miRNA World Congress’ and the ‘Epigenetics World Congress’, also organized by Select Biosciences, and registered delegates were given access to all three meetings.
Michael Metzker (Baylor College of Medicine, TX, USA) gave an overview of the next-generation sequencing technology platforms currently in use (standard Sanger sequencing being considered the ‘first’ generation), which he had also reviewed in a recent publication Citation[1]. Metzker was careful to state that there is currently no clear winner in the sequencing platform competition, but that different platforms may be better suited for different applications; for example, Illumina (CA, USA) and SOLiD (Life Technologies, CA, USA) do well in sequencing projects where a reference genome has already been established, while Roche’s 454 may be better suited for de novo sequencing and assembly. Metzker noted that there are more innovations on the horizon (i.e., next next generation), such as nanopore sequencing Citation[2], and that the cost of human-genome sequencing is decreasing dramatically, with one recent study from Complete Genomics, Inc. (CA, USA), indicating that the current costs for a whole-genome sequencing could be driven down to as low as US$4400 per individual, using DNA nanoarrays Citation[3]. Andy Carson (Korvis Automation, OR, USA) drew parallels between the current competition of the various gene-sequencing platforms versus the early days of the disk-drive industry, indicating that the eventual winners in sequencing will be those with the best technology that is delivered to market the quickest.
Innovative research into new and improved ways of sequencing is ongoing. Stephen Lindsay (Arizona State University, AZ, USA) presented his research on using electron tunneling to read the base composition of DNA. The basic idea of this approach is to drive unlabeled DNA through a very tiny hole, that is, a nanopore, so that the DNA has to pass through one base at a time; the tunneling junction has two electrodes, one to trap the DNA by hydrogen bonding to the backbone phosphates, and one to interact with the nucleotide base, giving a unique electronic readout for each base. Building on his previous work Citation[4], Lindsay stated that he can now identify all four bases in this manner. In principle, an array of such nanopores with electrodes would provide parallel readouts with increased accuracy and speed. This technology, once sufficiently developed, would have many advantages over current sequencing methodologies, including few sample preparation requirements, arbitrarily long read lengths and no inherent limitation to the four DNA bases (e.g., with possible applications to protein sequencing).
Next-generation sequencing has many applications beyond generating the base pair sequence of a genome. Next-generation sequencing can profile cells at various levels of biomolecular complexity, including gene transcription, DNA methylation, histone modifications, alternative splicing, small RNA profiling and DNA–protein interactions. While microarrays are currently in widespread use as a platform for molecular profiling, it is believed by many that, as its costs continue to decrease, sequencing will eventually supplant arrays as the preferred platform. At the conference, there were a number of talks showcasing examples of molecular profiling by sequencing.
Steven Jones (British Columbia Cancer Research Centre, BC, Canada) presented an intriguing example of personalized genomic medicine in practice. Jones’ group sequenced both DNA and RNA from the tumor of a patient with adenocarcinoma of the tongue, a rare disease for which the deregulated pathways have not been characterized. The patient had previously been treated with an EGF receptor (EGFR) inhibitor, but had ceased to respond. Copy-number analysis of the tumor revealed amplification of EGFR and deletion of PTEN, and DNA sequence analyses found mutations in TP53 and RB1, a combination that has been previously correlated with anti-EGFR therapy. Pathway analysis of the RNA profile suggested using a different drug to target the Ret oncogenic pathway, to which drug the patient initially responded. After the patient relapsed a second time, sequencing the anti-Ret therapy-resistant tumor revealed additional genomic mutations and alterations from the initial tumor.
Chad Creighton (Baylor College of Medicine, TX, USA) presented recently published work on profiling expression of miRNAs in ovarian cancer and the normal ovarian epithelium using sequencing. By integrating the miRNA data with corresponding data from gene-expression arrays, both the deregulated miRNAs and their predicted gene targets in cancer could be identified, and these targets could then be validated by manipulating the given miRNA in vitro. Molecular profiling of serous ovarian cancers (the most common subtype of the disease) led to the identification of the miRNA miR-31 as having antiproliferative effects in cell lines harboring defects in the p53 pathway Citation[5], and molecular profiling of clear-cell ovarian cancers (the most aggressive subtype) led to the discovery of miR-100 inhibiting mTOR signaling and enhanced sensitivity to rapamycin Citation[6]. Furthermore, a sizable number of novel miRNAs, not previously discovered using cloning or first-generation sequencing, were discovered in the cancers Citation[7].
Masako Suzuki (Albert Einstein College of Medicine, NY, USA) gave an overview of methods for DNA methylation profiling by sequencing. At the high-resolution end of the spectrum is the ‘MethylC-seq’ method from Lister et al., which can yield more than 25 million analyzable CpG sites, but requires a tremendous amount of sequencing (∼1.8 billion reads) Citation[8]. Methods with an acceptable trade-off between resolution and resources include ‘Hpall-tiny fragment enrichment by ligation-mediated PCR (HELP) tagging’ from Suzuki et al., which can yield approximately 1.8 million analyzable CpG sites and involves a more manageable number of sequence reads (5–10 million) Citation[9]. HELP tagging analyzes genomic fragments following restriction enzyme digestion. Sequences that flank sites digested by either a methylation-sensitive enzyme, HpaII or a methylation-insensitive enzyme, MspI, are read. Normalizing the HpaII counts by the MspI counts allows for quantification of the methylation.
Corey Nislow (University of Toronto, ON, Canada) shared his work on developing genome-wide dosage assays using next-generation sequencing. The aim of these assays is to discover which genes are essential for the growth of cells under various conditions. In the case of yeast studies, the assays are carried out on 6200 yeast strains, each strain with deletion of a single copy of a particular gene. In normal growth conditions, the heterozygous strain grows with the same kinetics as the wild-type yeast, but in abnormal conditions (e.g., during drug treatment), where the function of a given gene is critical, the strain with the single copy of the gene will not grow. As molecular barcodes are genetically linked to the gene deletion, all strains can be grown together, and then after the experimental treatment, the barcodes can be profiled to see which strains did not grow (identifying critical genes). The assays were initially conducted in parallel using microarrays, but can now be done using sequencing, as recently described Citation[10]. Nislow also shared his recent efforts in adapting the assay to study cancer cells.
Two talks presented sequence analyses of microorganisms. J Gregory Caporaso, from the Rob Knight Laboratory (University of Colorado at Boulder, CO, USA), presented a microbial community analysis, involving tens of millions of sequences, which captured known differences between microorganisms. Nick Bergman (National Biodefense Analysis and Countermeasures Center, MD, USA) presented the first analysis of a bacterial transcriptome by next-generation sequencing, which demonstrated how RNA structure and abundance can be defined at high resolution on a global level.
During the course of the conference, it became more and more apparent that a virtual explosion of new sequencing data is at hand. With the dramatic decrease in the time required and costs, next-generation sequencing makes possible the production of enormous volumes of cheap data, which could challenge the current infrastructure of information technology systems. A number of big science sequencing projects currently underway, including the 1000 Genomes Project, the Human Microbiome Project, the Personal Genome Project and The Cancer Genome Atlas, will lead to the sequencing of thousands of genomes. In the near future, the challenges and opportunities will shift from generating all this data to mining and interpreting them. Finally, as Robert Cook-Deegan (Duke University, NC, USA) discussed, another set of challenges in the legal realm of patent infringement may arise as sequence-based testing in the clinic becomes more common, since thousands of patents currently mention specific DNA sequences. One could argue in this case that broad gene patent claims should not be a barrier to furthering innovation in medicine.
Financial & competing interests disclosure
This work was supported in part by NIH grants P30 CA125123 and HL095382-02, and a Program Project Development Grant from the Ovarian Cancer Research Fund. The author has no other relevant affiliations or financial involvement with any organization or entity with a financial interest in or financial conflict with the subject matter or materials discussed in the manuscript apart from those disclosed.
No writing assistance was utilized in the production of this manuscript.
References
- Metzker M. Sequencing technologies – the next generation. Nat. Rev. Genet.11(1), 31–46 (2010).
- Branton D, Deamer D, Marziali A et al. The potential and challenges of nanopore sequencing. Nat. Biotechnol.26(10), 1146–1153 (2008).
- Drmanac R, Sparks A, Callow M et al. Human genome sequencing using unchained base reads on self-assembling DNA nanoarrays. Science327(5961), 78–81 (2010).
- He J, Lin L, Liu H et al. A hydrogen-bonded electron-tunneling circuit reads the base composition of unmodified DNA. Nanotechnology20(7), 075102 (2009).
- Creighton C, Fountain M, Yu Z et al. Molecular profiling uncovers a p53-associated role for microRNA-31 in inhibiting the proliferation of serous ovarian carcinomas and other cancers. Cancer Res.70(5), 1906–1915 (2010).
- Nagaraja A, Creighton C, Yu Z et al. A link between mir-100 and FRAP1/mTOR in clear cell ovarian cancer. Mol. Endocrinol.24(2), 447–463 (2010).
- Creighton C, Benham A, Zhu H et al. Discovery of novel microRNAs in female reproductive tract using next generation sequencing. PLoS ONE5(3), e9637 (2010).
- Lister R, Pelizzola M, Dowen R et al. Human DNA methylomes at base resolution show widespread epigenomic differences. Nature462(7271), 315–322 (2009).
- Suzuki M, Jing Q, Lia D et al. Optimized design and data analysis of tag-based cytosine methylation assays. Genome Biol.11(4), R36 (2010).
- Smith A, Heisler L, St Onge R et al. Highly-multiplexed barcode sequencing: an efficient method for parallel analysis of pooled samples. Nucleic Acids Res.38(13), e142 (2010).