Bio 363 Human Genetics Spring 2013
Autism study sheet
Back to syllabus

Background web reading:
What are characteristics of individuals on the autism spectrum?
What other conditions tend to co-occur with autism?
What is the prevalence of autism, and how has the prevalence changed? Why might it have changed?
What do you think of the Autism Speaks website’s assertion that parental age is an environmental, non-genetic risk factor for autism? (come back to this question after reading the review papers)
What are SNPs? What is linkage disequilibrium (LD)? How do these two things relate to the HapMap? How is the HapMap used for genome wide association studies (GWAS)

Visscher et al., 2012 (GWAS review)
This paper has a somewhat defensive tone—why?
Does a GWAS determine a causal variant for a disorder?
Have GWAS’s led to any useful functional information for any disorder? (And therefore is the criticism/cynicism justified?)
Focus on page 17, middle of right column, section on differentiating between two explanations for observed associations: what does this mean?
After you read the review paper about de novo mutations and disease, consider how that phenomenon might affect GWAS’s.

Veltman and Brunner, 2012 (de novo mutation review)
Where and when do de novo mutations associated with sporadic disease occur?
What are different types and rates of occurrence for de novo mutations?
In a general sense, how do those rates plus other factors translate to the “burden of disease that is due to de novo mutations”?
Why are most new single nucleotide variants (SNVs) from the paternal side?
What are the main technical challenges of detecting de novo mutations?
What are the challenges and approaches in determining whether a de novo mutation is actually pathogenic?
How should genetic counselors advise parents whose first kid had a de novo mutation?
What sort of prenatal screening might become the norm?

Weiss et al., 2009  
Note that this paper has some true linkage analysis as well as some GWAS-type approaches.
Don’t worry if you don’t understand all the statistical tests, etc. Just get a strong intuitive sense of the findings.
What were some of the quality-control measures that they took?
Figure 1: Which regions of the genome show linkage with autism?
Table 1: Look up information on the transmission disequilibrium test (TDT) and figure out how it is different from traditional linkage analysis. What data in this table are only from TDT? What additional approach was used to strengthen the data?
What happened when the researchers tried to replicate the results with more families?
What is the one genomic region of strong interest and what genes are there?
Figure 2: What evidence is there that SEMA5A might be involved in autism?
What happened when previously-discovered genes of interest were examined for association with autism in this data set?
How do the authors explain why their linkage results and their association results don’t correspond? (top left of p. 805)

Sanders et al., 2012  
Look up next-generation sequencing techniques in general, and Illumina sequencing in particular. (Very different from old-fashioned Sanger sequencing!)
Why was it important to have samples from families with two parents and both an unaffected and affected child?
What sort of quality control measures ensured that sequencing mistakes were not misidentified as true mutations?
Figure 1: What is different and what is the same between probands (affected children) and their unaffected siblings?
Figure 2: Without trying to understand the details of the modeling, get a sense of why they had to figure out the false discovery rate and how this work validated the idea of looking for two independent mutations in the same gene.
What gene was identified in this way, and why is it a good candidate?

O'Roak et al., 2012  (Science 338:1619)
How is the overall approach in this paper different from that in the previous main paper?
What is the MIP strategy?
What criterion focused attention on a subset of genes?
How did the researchers determine (roughly) that they detected more mutations in the genes in question than one would expect randomly?
Why did the researchers examine head circumference in some of the probands?
Briefly look through the huge amounts of supplementary materials at the links here and see if there's anything worth bringing up in our discussion.

There are a few papers in the "extra" folder on the public server. Just glance at abstracts and conclusions. What techniques were used? How does it all fit together?
The Cell paper by Michaelson et al. has one of those symbols you can scan with your phone to go directly to a web site. It takes you to this video: