SoylentNews is people
SoylentNews
I am part of an international team of researchers who have been exploring the capabilities of the MinION Sequencer, the MinION Analysis and Reference Consortium (MARC). Our first paper on this exploration has just been published in f1000 Research. Five separate labs carried out four sequencing runs each of the same strain of E. coli, and a few more labs helped to analyse the results. If you're interested in seeing what this technology is capable of (or at least, what it was capable of about 6 months ago), check out the paper here, or download the data here.
The Oxford Nanopore MinION is a small DNA sequencer that plugs into the USB port of a laptop and sequences DNA by measuring changes in an electric current as the sequence is passed through one of 4096 pores in the sequencing device. These electrical signals are combined into events that describe the movement of a single base, and the events are then base-called to generate DNA sequences.
The MinION sequencer is almost entirely electronic, stripping away everything that makes existing DNA sequencing technologies big, heavy, slow and expensive. This has meant that the MinION is uniquely able to be used in remote areas where other sequencers just can't reach: sequencing Ebola on-site in Africa, sequencing the DNA of small frogs in the Amazon rainforest, and more recently sequencing DNA in NASA's vomit comet.
Previously: The MinION - Genome Sequencing in a Handheld Device
(Score: 4, Interesting) by gringer on Monday October 19 2015, @10:26PM
What is it called when multiple cells are sequenced to reduce errors?
Overlap-consensus assembly, if you're combining multiple reads.
What kind of accuracy does MinION get?
Per-base accuracy is about 88%, with no dependence on read length, which only starts to look impressive compared to Illumina reads when you have reads that are over 1kb in length. With overlap consensus and careful library preparation, one person has achieved 99.8% accuracy [github.io]. That's still a long way from the "6 nines" accuracy of consensus PacBio sequencing, but it's a start.
Do you simply use it for longer in order to get more accurate results?
Yes. That works as long as you have more than one sequence to play around with. Here's an example of how a customs agent might use it from a prepared library:
[what am I looking at here]
[is it the "safe" version of that thing]
[did it come from Steve's farm, or Jenny's from just down the road?]
More answers to follow a bit later....
Ask me about Sequencing DNA in front of Linus Torvalds [youtube.com]
(Score: 3, Informative) by gringer on Tuesday October 20 2015, @03:43AM
That works as long as you have more than one sequence to play around with
I should probably expand on that statement. Everything that gets loaded into the MinION sequencer will be read at most once. A protein is attached to the sequence which allows it to combine with the nanopore and process through the hole, and that protein is ejected at the end of the sequencing process. If you have a single gigabase-size sequence available, none of the reads that come out of the sequencer will overlap unless there is duplication in the genome.
It is apparently possible to pull the sequenced library back out of the device, re-prepare it, and then run it through again, but I'm not aware of anyone in the MAP community who has experimented with that.
Ask me about Sequencing DNA in front of Linus Torvalds [youtube.com]