One rarely hears mention of deep sequencing without “more” being dropped into the sentence. More coverage, more depth,…more sample? That’s right. As valuable as next generation sequencing platforms have become, those library construction protocols demand a fair amount of sample. Often more than can be coaxed out of a biopsy, or other small cell population. Fortunately, researchers at the Broad Institute and Helicos BioSciences recently teamed up to develop a technique that ditches the library and cuts right to the chase with direct sequencing, enabling analysis of as little as 50 pg of DNA, the equivalent of only about 25,000 cells.
The new method allows researchers to bypass many of the processing steps associated with conventional ChIP-Seq, such as adaptor ligation, DNA size selection, and PCR. After ChIP, single molecules of DNA are sequenced directly with the Helicos HeliScope Genetic Analysis platform. When the team used the technique to analyze DNA bound by posttranslationally modified histones or by the protein CTCF, they found that:
- There was good agreement between data acquired from the Helicos and Illumina Genome Analyzer sequencing platforms.
- Because PCR is not required, the new method showed less GC-content bias than a conventional ChIP-Seq procedure.
- The technique more accurately estimated the size of the DNA region protected by bound proteins.
- 100-fold less starting material was required for the direct sequencing approach than for conventional ChIP-Seq.
Cutting corners won’t deliver in every scenario, however. Since direct sequencing yields relatively short read lengths (25-55 bases), it probably won’t be the “go to” method for applications that require deeper genome coverage.
“For samples that are easy to obtain (e.g., cell lines) and for modifications that are technically easy to enrich for (e.g., histone modifications), it will take time before direct sequencing will replace next-gen sequencing,” says Alan Goren, lead author of the study. “Nevertheless, because this technology provides opportunities to generate epigenomic maps from primary cells and other hard-to-obtain cell populations, I believe it will be highly used. The field really needs ways to generate data from non-tissue-culture samples.”
See what the big deal is at Nature Methods, November 2009.