In this short (11 minute) TEDxBoston lecture Dr. Richard Resnick gives the hard sell on widespread genome sequencing.
A few quick pointers or thoughts from my viewing of it:
- The ‘software’ he refers to are those developed by people in my research field, bioinformatics or computational biology. This software is an integral part of sequencing a genome using current methods.* The software that assembles the smaller sequences into genome sequences is doing what is called—surprise, surprise—sequence assembly. Likewise algorithms are used to locate differences between different genomes (like the comparison of normal and cancerous cells he shows), and software applications process that data at every step.
- A ‘gotcha’ with the claim of cheap genome sequencing is that while the cost of the biochemistry of getting the raw genome sequence data is falling, is the same true for extracting information from it? (Readers with subscriber access might which to read The real cost of sequencing: higher than you think! – Genome Biology 12:125.)
- His talk a hard sell in that the medical examples he used ‘worked’ in that some positive action could be take as a consequence of the test. A limiting step is the knowledge needed to make good what you have learned from the genome sequence. I’m not saying a genome sequence is worthless, here, but that having got it, you still need to know how to use the information you’ve learnt.
Footnotes
* There is work towards methods to sequence that attempt to sequence very long pieces of DNA at once; if reliable and cost-effective these potentially could displace the need for sophisticated algorithms to assemble genomes. I don’t particularly like that he refers to it as ‘software’ as there is a distinction—to me, at least—between application software, statistical analysis, algorithms and so on. But this is a TED lecture, I guess…
(You can read elsewhere on this blog for my distinguishing bioinformatics and computational biology.)
Other articles on Code for life:
Boney lumps, linkage analysis and whole genome sequencing
Genetic tests and personalised medicine
Appeals court concludes that Myriad can patent BRCA genes
Haemophilia – towards a cure using genetic engineering
Grant, you may be interested in the followings if you haven’t come across them (or have develop those algorithms) already.
- Applications of Signal Processing Techniques to Bioinformatics, Genomics, and Proteomics
- Digital Signal Processing for DNA Sequence Analysis
Those documents above don’t cover Wavelet, but I’ve seen a few publications on the net in the use of Wavelet in Bioinformatics and DNA sequencing analysis. IMO, the best open source Wavelet software available on the net is the WaveLab from Stanford, which is written in Matlab. I ported part of WaveLab into Java for an industrial image classification application project that I was involved in.