ASHG 2010 is now over, and I am back on Albion. Either me, or Daniel, or both (or, indeed, any of the other GNZers) will have personal genomics roundup over at Genomes Unzipped sometime this week.
For the last of these posts, however, I thought I would just report on the entirety of the Exome Sequencing session on the final day of the conference. I loved this session for the diversity, the number of different projects that are using exome sequencing to address old questions. It shows how much biology is tech-limited: the moment a powerful new technology becomes available at a low price it is used in every field by a flood of researchers who have been waiting for exactly this sort of data.
Other than that, there wasn’t an overall theme to the session (or to this blog post), other than Exome Sequencing Is Cool.
Another interesting day at ASHG so far (and not over yet). As with last year, genotype imputation (using reference sets to infer the genotype of untyped variants in your samples) has been a pretty major subject of the meeting. In particular, the idea of using large sequencing refernce sets like the 1000 Genomes Project to infer lower frequency variation in existing Genome-Wide Association Study datasets has been raising people’s hopes for accessing new types of variation “for free” (i.e. without having to regenotype samples).
Getting at Low-Frequency Variation
The “Genome-Wide Association Studies and Imputation” session started off with Vasyl Pihur’s somewhat provocatively titled talk “Neither common nor rare variation can explain much of phenotypic variation”. The point he was making (and confirmed with some model fitting to existing datasets) was that it is hard for very rare variation to explain much heritability, because so few people carry any particular variant, and very common variation has still left much heritability unexplained, so our best bet for filling in “missing heritability” is varients of intermediate frequency, the neither-common-nor-rare “low frequency” band that lives between 0.5% and 5%.
All the ASHG talks that I have had to do analysis for have now been given, so today I’ve managed to dedicate my full attention to the sessions. Also a good day for tweeting; I managed to live-tweet quite a few talks on @lukejostins, and the #ASHG2010 hashtag has been totally rammed.
Larry Parnell over a Varigenome has been putting his ASHG notes up, if you are still hungry for details. Daniel Macarthur has promised a post on the “Identifiability in the Era of Genome-Scale Research” session for Genomes Unzipped, and I saw him getting pretty worked up about Jim Evan’s talk on his twitter feed, so hopefully we’ll see something from him as soon as he’s done being dead of plague.
Two sessions today, “Statistical Analysis of Human Sequence Variation” and “Finding High-Risk Susceptibility Gene Variants”, seem to encapsulate the cutting-edge of disease gene association, and illuminate where disease genetics is heading in the immediate future.
Unfortunately, I’ve had a bit of a distracted day today; some analysis that is being presented tomorrow failed, and as a result I dropped off the radar somewhat trying to put it right.
I sat in the “Lessons from High-Throughput Sequencing” session, and picked up bits of it, but was generally distracted. Zam Iqbal presented some very impressive work on assembling genome sequence as diploid individuals, which will be extremely important in the future. The main reason for this is that it allows accurate HLA typing from whole-genome sequencing; HLA typing costs hundreds of dollars, so getting it for free as part of the genome is a major win for personal genomics.
Low-Key Personal Genomics
An interesting, though not new, story came in the form of back-to-back talks by Euan Ashley and Russ Altman on the disease and pharmacogenomic work on Steve Quake’s genome.
As if no time has passed at all since the sunny shores and lost laptops of the American Society of Genetics 2009 meeting, ASHG2010 has rolled around, this time in Washington DC. As always, I’m going to be trying to write a few thoughts on the conference every day, though this year it may be split between here and Genomes Unzipped.
I’ll also be semi-live-tweeting (wifi coverage is patchy), so you can get up-to-the-minute details of all the talks on my twitter feed (@lukejostins), or from other tweeps via the hashtag #ASHG2010.
Epistasis and Missing Heritability
As Daniel observed on Twitter, I very nearly had a heart attack when Eric Lander, in his Distinguished Speaker’s talk about the Human Genome Project, said that the “missing heritability” is probably all down to Epistasis (i.e. interactions between variants). His argument was that GWAS had low power to detect gene-gene interactions, and therefore there could be lots hanging around that count account for the unexplained variance.
This is a fallacy, and a big one. Continue reading