This year I'm writing these predictions at PAG, the Plant and Animal Genome Conference, where everything under the sun is discussed under the sun in California. The somewhat relaxed atmosphere before and after the conference sessions is not particularly conducive to forward thinking, but I'll give it a go.
1. Bacterial metagenomics. Next Big Thing. Consumer goods and food companies have known about this field for years, but this year we'll see big leaps forward in pharmaceutical applications of microbiome research, and also in soil for plants.
2. Nanopore sequencing. Will they, won't they? Who knows. My guess is as good as anyone's so I'll repeat last year's prediction that this will be the year. I really hope it will be this time.
3. The first pan-genome will be published. We have to stop relying on the minutely-detailed annotation of a single individual and instead look at understanding the general pattern of an entire population - the pan-genome - in order to better pick out variant signals that are actually of interest. Think about a tomato, for instance - there are many varieties, all of which are identifiably tomatoes, but with huge genetic differences between them. A tomato pan-genome comprising the sum total of all known tomato varieties would identify what it really means to be a tomato and help narrow down the target areas for varitaion detection that are genuinely unique to the feature studied rather than being an artifact of that particular strain.
4. Metadata matters. Eagle surely aren't the only one working on a product in this field, so expect to see at least one competitor make the news with their brand new alternative solution (nowhere near as good as ours, naturally) before the year is out.
5. Where there is metadata there must be order, which means standards. We use the ISA model, and so do Nature Publishing Group and the BGI. With backing like that we hope it will become the de-facto method of organising metadata. I predict that this year we'll see a major institute - maybe the EBI or Broad, as randomly-selected examples - announce that all their publicly released data will be annotated in ISA format.
6. AstraZeneca are moving to Cambridge from Manchester. We wish our friends there the best of luck for the move, but given the current state of its pipeline will AZ become the target of a takeover before it manages to break ground at its new campus? This prediction says (and hopes) not.
7. If AZ don't get taken over this year, someone else in pharma will. (Is that vague enough?)
8. The trend towards more natural food products and an emphasis on tracing the origin and quality of ingredients will see an increasing number of food companies start to put real effort into using genomic techniques to improve source crops (better selective breeding, not genetic modification) and monitor the provenance of their products whilst in transit. The UK horsemeat scandal is a perfect example, but cocoa and coffee companies and others reliant on promoting their products as fair-trade or single-origin face an increasing need to improve their systems for enforcing and backing up these claims. Whether this will be driven by the purchasers - supermarkets, for instance - or the brands themselves remains to be seen. My prediction is that we'll see at least one major new consortium set up between food companies to advance pre-competitive research in this area for a commercially important and widely used ingredient. (One already exists for oats, for example.)
9. Speaking of horsemeat, it transpired that whilst checks were taking place, nobody found the contamination because nobody was looking specifically for horse. The problem with current techniques is that they assume a limited range of possible contaminants and only test for those, as opposed to comparing the sample to all known organisms to find out what is in it, expected or otherwise. Horse testing may well have been stepped up now, but it's a fair bet that nobody is testing for buffalo, ox, or other meats similar to beef. And then there's the fishing industry where passing off processed products from a cheaper fish as a more expensive variety is (apparently) rampant. To solve this requires a low-cost rapid-results test that performs comparisons between the genome of the sample and the genome of all known animals and contaminants (bacteria, etc.), which is asking a huge amount. But, it is possible, and whilst we may not see widespread application of a commercially-viable tool this year, I think we'll see the first attempt released by a university somewhere.
10. Fish! During 2012, farmed fish outproduced farmed beef for the first time globally, and in 2013 was expected to outperform wild fish consumption too. But where are the fish genomes? There's a handful such as lamprey, fugu, or zebrafish which were sequenced for other purposes, but virtually nothing complete in the public domain related to edible fish. Why are we paying so little attention to edible fish when the market for it is so much larger than the well-understood cattle, sheep, and pig genomes? This year I hope will see an upsurge in the number of good-quality well-annotated edible fish genomes published. At least the salmon if we're going to be Western-centric, to join the existing cod genome, but seabass or tuna wouldn't be a bad start either.
Well, we got to ten predictions just about at the point where I ran out of ideas, so that is a Good Thing. They're a bit different to last year's in that they're more big-picture and less about the technical detail, but the bioinformatics industry does somehow feel a lot more stable and a bit more mature (in terms of changes to tools, tecnhiques, and development of new research areas) than it did a year ago, so this is reflected here. I look forward to getting everything completely wrong again, as usual, except the ones where I hope things will go well, which I look forward to getting right...