December 2016: Scientists from the Wellcome Trust Centre for Human Genetics (WTCHG) have used Oxford Nanopore Technologies MinION sequencing devices to sequence two human DNA samples, in an exploration of the capabilities of the pocket-sized, USB-powered sequencing devices.
Oxford Nanopore Technologies has developed sequencing and sensing technologies using nanopores – tiny-sized holes in membranes through which molecules such as long strands of DNA can pass. The DNA is detected, and can be sequenced by measuring the fluctuations in ionic current as each nucleotide, or letter, of the DNA strand passes through the pore. The Oxford Nanopore Technologies MinION device contains hundreds of pores, each connected to electronics, in a device only a few centimetres long that is powered by and connected to a laptop computer.
The WTCHG has worked with scientists from Genomics PLC, an Oxford University spinout whose expertise is in the analysis of complex genomic datasets, to sequence and analyse the genomes of two human individuals. The first of these samples, known as ‘NA12878’, is DNA from a cell line derived from a donor, whose genome has been studied many times before using other technologies and provides a benchmark for measuring the performance of the Nanopore approach. The second, known as clinical sample ‘X’, will provide insights into the potential value of Nanopore sequencing for understanding individual cases of human disease. We have used eight MinIONs over the course of about two weeks to generate all the data presented at Oxford Nanopore’s Nanopore Community Meeting in New York on 1 December 2016 and included in an initial data release.
Compared with existing methods for sequencing whole human genomes, which produce data in many millions of small fragments that need to be matched to a reference genome sequence, Nanopore sequencers produce longer chunks of data that provide better information about the structure of the genome but also contain more sequence errors. The longer reads of Nanopore data should allow us to directly detect genomic changes that may be important in disease, such as the re-arrangement of genes, which currently may be impossible without further investigation. Where a single gene in a sample contains more than one change that might cause disease, long reads can also help to identify which of the two copies contains the change; information that may be important in understanding the disease. At the same time, the higher rate of errors in individual reads provides its own challenges.
These datasets are currently being analysed for publication.