Impact Of Long-Read, Single Molecule Sequencing

Question

[Damian Kao]

The whole community is buzzing with speculation about the Oxford Nanopore after the presentation on Friday at AGBT. Here is a interview with Clive Brown on the technology and all the changes that might come with it.

In the interview, Clive Brown seems to emphasize the need for different computing hardware architecture where raw data processing is done in real time as the data streams out. He also touches on new file formats and perhaps using a HDF5 data model for storing information.

If this technology really performs as well as the claim and they increase the poor accuracy (currently 96%), how will this impact the field? We can obviously see what problems having long reads and high coverage will solve, but what new problems will it produce? Will current generation short-read algorithms become obsolete?

Answer 1

[Istvan Albert]

Frankly this sounds too good to be true - and in general that means just that.

Either way, no matter what technology comes around I expect the cost of sequencing to drop while the cost of data analysis to rise.

Answer 2

[Zev.Kronenberg]

A curiosity of mine is what these data actually looks like? What is the distribution of read lengths? Quality? Are any data publicly available?

I know that longer reads will result in better phasing which will be nice. I suppose estimates of LD will also get beter with longer contiguous reads.

Answer 3

[Ryan D]

I think it would behoove Oxford Nanopore to release some of the raw data coming off the sequencers to those working in the field. Assembling would ostensibly be easier, right? Will the software to do so be the same? Will the workflow be the same or different?

More importantly, will this allow me to sequence from my computer terminal and keep DNA samples in the little mini-fridge I keep stocked with beer for after-hours?

Good-bye core labs and 3-month waits! Hello Dilbert-land sequencing!