Researchers at Cold Spring Harbor Lab recently developed a sequence analysis strategy they call “DNA Sudoku”.
Cute name, right? I have to admit, it was the name that first caught my eye. But the Sudoko method is much more than just a cute name, it is a significant advance in multiplex sequencing.
Multiplex sequencing, or multiplexing, is a fairly recent development made possible by next-generation sequencers. Multiplexing relies on tagging DNA samples with short oligonucleotide “barcodes” which allow for the sequencing of many different samples at once. The advent of multiplexing was a substantial improvement over the previous techniques which could only sequence one DNA sample at a time. However, construction of a library of unique barcodes is expensive in terms of both time and money.
The authors of DNA Sudoku – harnessing high-throughput sequencing for multiplexed specimen analysis, report that they:
…have developed a novel multiplexing framework in which the identity of individual specimens is not encoded directly by an appended barcode sequence. Instead, specimen identities are encoded by positional information created by an optimized series of pooling schema. In this manner, more than 100,000 different samples can be analyzed using only a few hundred barcodes
Using the positional information of pooled samples was an ingenious innovation and because it decreases the number of barcodes required, it significantly decreases the costs of multiplex sequencing. The use of positional information is also where the method gets its name.
While each individual pooling pattern might yield multiple solutions to the link between sequence and specimen source, the combination of all pooling patterns would provide sufficient constraints that only a single, high-confidence solution would emerge for the vast majority of samples. Many elements of this approach were reminiscent of seeking the solution to a Sudoku puzzle, which led us to dub this strategy ‘‘DNA Sudoku.’’
The pooling strategy relies on number theory and is based on the Chinese Remainder Theorem. I didn’t delve too deeply into those ideas – I mostly took the authors’ word for it. But if you want to look into the details behind the background on the number theory, the authors’ suggest Du and Hwang 2006a for review. The important thing is that it works. And it does work. Their correct decoding rate was 98.2%, as good as or better than most current barcode methods.
The low cost and high accuracy of the Suduko method dramatically lowers the barriers to genotyping and may lead to major advances in human health research. I am very interested to see how quickly this new method is adopted and just how much impact it ultimately has.
0 Responses to “DNA Suduko”