Give Up Blog - for scientists like these!


You hid my archives, didn't you Steve!


Maps and Figures

"Hitler or Coulter?" Quiz
Map1 - Teen Pregnancy
Map2 - Incarceration
Map3 - Homicide Rates
Map4 - Drop-out Rates
Map5 - Bankruptcy Rates
Map6 - Driving Distances
Map7 - Energy Use
Map8 - Gonorrhea!
Map9 - Tax Burden
Map10 - State GDP
Map11 - DHS funding
Map12 - Adult Illiteracy.
Map13 - Abortion Bans:
Map14 - ER Quality
Map15 - Hospital Quality
Map16 - Coal Burners
Map 17 - Infant Mortality
Map 18 - Toxic Waste
Map 19 - Obesity
Map 20 - Poverty
Map 21 - Occupational safety
Map 22 - Traffic deaths
Map 23 - Divorce
Figure 1 - Wages vs Right to work
Figure 2 - Unemployment vs Right to work
Give Up Shopping guide

Email



link to xml feed Subscribe with Bloglines

Google


www

giveupblog

Monday, February 19, 2007

Autism, SNP arrays and Neurexin 1
Hail genomic medicine! Researchers have used massive genetic screens to identify a gene with a major link to autism in Nature Genetics AOP (BBC article here).

Here's the abstract:

Autism spectrum disorders (ASDs) are common, heritable neurodevelopmental conditions. The genetic architecture of ASDs is complex, requiring large samples to overcome heterogeneity. Here we broaden coverage and sample size relative to other studies of ASDs by using Affymetrix 10K SNP arrays and 1,168 families with at least two affected individuals, performing the largest linkage scan to date while also analyzing copy number variation in these families. Linkage and copy number variation analyses implicate chromosome 11p12–p13 and neurexins, respectively, among other candidate loci. Neurexins team with previously implicated neuroligins for glutamatergic synaptogenesis, highlighting glutamate-related genes as promising candidates for contributing to ASDs.


Autism is complicated though and even in twin studies there isn't a 100% correlation (it's 60-90%) suggesting that even if you did have one or more genes that contribute to the disorder you're still not guaranteed to get it (this is called incomplete penetrance of a trait). This suggests environmental factors, early childhood events, something in gestation etc., that might be additionally required for the disorder to present.

Autism actually has an interesting history in terms of its suspected causes. First described by Leo Kanner, the suspected causes followed a pretty predictable course for psychiatric disorders of children. What I mean is that of course, first the parents are blamed for the disorder (specifically the mother). This was the fault of Bruno Bettelheim who coined the term "refridgerator mother" to describe the emotionally distant mothers who must have caused the problem. Luckily this view eventually fell into disfavor, but there are still plenty of idiotic theories as to the causes, like vaccines. I won't say mercury, because after mercury was removed from vaccines the rate of autism continued to increase, and the anti-vaccination wackos have persisted in attacking vaccination as a cause of autism. Autism has also had a major uptick in frequency in the last few decades, but this is more likely to be due to increased ability to recognize the full spectrum of autistic disorders and inclusion of kids with "autistic traits" into the definition. Right now the current frequency of autistic disorders is estimated at about 1:150.

Now, the researchers of The Autism Genome Project Consortium have made a major advance in understanding autism through the discovery of genes frequently linked to the disorder. They did this by analyzing DNA from about 1400 families in an attempt to statistically link single-nucleotide polymorphisms (SNPs, pronounced snips, meaning a single nucleotide deviating from norm in the genome) to the autism spectrum disorder. They do this using SNP gene arrays made by a company called Affymetrix, who make most of the popular "arrays" for rapid analysis of thousands of genes by their level of expression or by analysis of polymorphisms. These arrays consist of a solid surface with various known DNA sequences that are spotted into specific positions. The samples of genomic DNA are then loaded onto these arrays (and hybridize by Watson-Crick base pairing) and then are probed for the presence of these single base mutations. This allows researchers to evaluate the entire genome of an individual very rapidly. It also generates an enormous amount of data, an enormous amount of cost, and an enormous amount of time interpreting the output. So you can imagine with 1400 families being tested, this is a truly phenomenal undertaking.

Their results? They were able to link many copy number variations (CNVs - repetitions or deletions of between 10,000 - 5 million bp of DNA on chromosomes) in various regions of the chromosomes, with the presence of autism in individuals. The full set of CNVs is quite large and it's not clear which genes they contain and increase the copy number of, but one in particular set off their alarms when they detected a deletion of an exon in the neurexin 1 gene that was associated with autism in one pair of individuals. While this isn't a completely solid finding since it was such a small sample, the interaction of neurexins with other genes implicated in autism disorders and the presence of this mutation in two affected individuals (and absolutely no controls) is suggestive of an important piece of information in the autism puzzle.

This is pretty amazing stuff, and while massive informatics like this isn't my specialty, I can appreciate what an amazing feat this is.

Labels: ,