This web page was produced as an assignment for an undergraduate course at Davidson College.

image: iStockPhoto

D. Benjamin Riffe's Genomics Web Page

Home Schizophrenia Paper Review #1 Paper Review #2

Writing for scientific and academic journals differentiates itself from identically themed writing for popular newspapers and magazines.  To be sure, those writing to and within the scientific community and academia write to a different audience with general scientific literacy and acceptance of jargon.  In contrast, popular journals, to increase readership and interest in articles, tend to use less scientific jargon and less complex language.  The most surprising part of Kua et al’s (2004) study, however, was the author of popular journal pieces’ tendency to write and report something different than the scientific journal had reported.  Kua, et al (2004) suggests that popular journal authors work as an “intermediary” between the scientific writing and research and their standard readers.  Kua et al (2004) also writes that the journalist should serve as “watchdog” of the scientific community and put the findings in sociocultural context.  Lastly, and perhaps done least often, the journalist should serve as “tool-giver” and provide the context for the science in the current body of science so readers can make useful and applicable inferences and ask questions.  Within this framework, we will compare a scientific article out of the peer reviewed journal Nature by Stefansson et al (2008) and an article about the study from the New York Times “Gene-Hunters Find Hope and Hurdles in Schizophrenia Studies” by Wade (2008).  This provides an opportunity to compare two real examples of scientific reporting, as well as an opportunity to explore the role of genomics in understanding schizophrenia. 

Large recurrent microdeletions associated with schizophrenia
in Nature

Drawing on previous research, which showed that rare copy number variations (CNVs) had been associated with mental retardation and bipolar disorder, the researchers looked genome wide for new CNVs in patients with schizophrenia.  This method of sequencing genomes to look for rare variants and deletions stems from the hypothesis that rare variants make up the current autism and mental retardation because risk alleles are selected against, so there must be several newer risk alleles.

A two-phase large-scale sampling was used. In the first phase, regions of the genome that replicate at a high rate (>1 in 10,000 divisions) were enriched in order to look for de novo CNVs.  If particular CNVs had a low frequency in the population, then the CNV is under negative selection pressure and a likely candidate for a risk CNVs.  The methods to find new CNVs are discussed, using a 9,878 parent and children combinations without schizophrenia and differentiating the CNVs into variations with low copy repeats (LCRs) contiguous to the CNVs, and CNVs found in one sample.  These 66 de novo CNVs were then tested against 33,250 controls from the SGENE consortium and 1,433 patients with schizophrenia.  On chromosomal positions 1q21.1, 15q11.2, and 15q13.3, three CNVs were found to associate with schizophrenia and psychosis nominally (Table 1) and significantly (Table 2).

The article the emphasizes that the standard practice of classifying mental diseases by their similarity to others, and that other mental illnesses, such as autism, could be associated with CNVs at chromosomal regions 15q13.3.

The paper continues to describe the variation in the length of the deletions at each site and discusses many genes found at those regions. The 1q21.1 deletion is 1.34 Mb and has many genes, including the GJA8 associated with schizophrenia and the improbability that other rearrangements or translocation are possible with this site. 

The 15q11.2 470 kb deletion contains several genes associated with a subgroup of Prader-Willi syndrome, Angelman syndrome, which have aspects of obsessive compulsive disorder and autism.  The schizophrenia phenotype CYFIP1 gene is affected by a mutated fragile X mental retardation protein (FMRP) by reducing gene expression of CYFIP1. 

The 15q13.3 deletion prevents the expression of several genes including the α7 nicotinic receptor gene (CHRNA7), potentially linked to schizophrenia and mental retardation. 

Structure of fragile X mental retardation protein

The paper concludes with the frequency of mutations lack of sex bias.  No patents with the three deletions were mentally retarded.  Additionally, no LCRs, which flanked the CNVs, could explain the variety of the deletions through recombination.  The authors ultimately suggest that future studies look into the CNVs associated with schizophrenia and autism are classic schizophrenia or microdeletion syndromes, as well the effect different CNVs have on drug treatment and symptomology. 

Methods

The authors discuss the methodology of their paper.  This included posting their approval by the National Bioethics Committees in the various countries where the data were taken and posting the consent of all 4,718 patients.  The authors also posted the sources of all the data, including the SGENE samples, CNV data from the HumanHap300 chip as well as additional resources and labs.  The researches posted explained the statistical methods used, which included the significance threshold of de novo genes associating with schizophrenia, that the P-values were two-sided in Tables 1 and 2, and the use of the Cochran-Mantel-Haenszel test for schizophrenia and CNV correlation. 

In the online version of the paper, the authors published a more complete methods, which included how a de novo CNV analysis is preformed and the the methods of the dosage measurement of Taqman assays, which were used in the Danish and Chinese samples of Table 2. 

Gene-Hunters Find Hope and Hurdles in Schizophrenia Studies
in The New York Times

This article published in The New York Times begins by exclaiming that the search for schizophrenia genes was the largest scale ever undertaken.  The findings of the research are then written in the second sentence and that there are rare variants for schizophrenia, although they only make up a few of the cases.  The article then discusses the actual results of the study, which showed that three variants in particular with large deletions of DNA from specific sites.  Interestingly, the article then puts the context of a recent finding by researchers from the University of Washington, which had found that rare deletions and duplications are associated with schizophrenia.  The hypothesis is that natural selection selects against schizophrenia genes because of the lower frequency of the schizophrenia deletions and duplications in relation to the non-schizophrenia deletions and duplications.  New cases of schizophrenia are caused by new mutations in the population.  Interviews with various professors working on the study revealed that many, such as Dr. Goldstein, believe that schizophrenia and other mental diseases are caused by new mutations and that they will not find common variations among all patients.

This article discusses the large number of studies that have been looking for common variants.  New developments in genome sequencing, specifically the hapmap, helped the researches look at the genome and variations with no expectations.  The article finishes by discussing the implication of no common variant causing schizophrenia on treatment.  Looking for similar up- or down-regulated pathways among patients should reveal similar treatment methods.  This finding should, however, change how schizophrenia is diagnosed. 

Comparison using the paradigm of Kua et al (2004)

Lack of Discussion of the Methodology

The New York Times article does not discuss the methods of the study.  There is a brief comment about the size of the study, but no real numbers are given.  Compared to the article in Nature, numbers of patients involved, the diversity of the types of mental illness the patients had, the sources and labs of the data, techniques, statistical analyses, and even a basic and rudimentary explanation of the experiment were absent. 

Change of Emphasis

The New York Times article recapitulates much of the emphasis of the original paper, which includes significantly associating certain microdeletions and CNVs with schizophrenia, while explaining the frustration of not having common variations to associate with schizophrenia. 
Intermediary

As an intermediary between the researchers, the New York Times translates the science and jargon into easy to understand language.  Results and the significance of those results are well written, capturing the nuances of progress with identifying markers of schizophrenia and the absence of simple Mendelian genetics.  Complex concepts of new mutations as the cause of schizophrenia are handled adequately. 

Watchdog

As the “watchdog” looking at the social and ethical implications of the research, the New York Times asks the questions of the researchers of this impact on treatment and diagnosis.  The New York Times article goes slightly further than the Nature article, explaining that treatment will not necessarily be more difficult, and future research using the mechanisms that cause schizophrenia to treat the patients.

Tool-giver

From a tool giving perspective, the New York Times article does not include a basic explanation of the actual research, or the reasoning behind the experiments.  There are no explanations of the number of people tested, the statistical tests run, the scope of this particular research project.  Readers of the New York Times do not have the opportunity to understand the methodology or think critically about how the science works.  The New York Times does attempt to put this research in context, and does place it in the recent findings of the University of Washington on rare deletions and duplications three months earlier.  The article mentions the thousands of studies on the genetics and genomics of schizophrenia, but fails to put the particular research in the CNVs related to autism and mental retardation.  The New York Times article also fails to explain well that these deletions put one at risk of developing schizophrenia rather than cause it in a majority.  Statistical analysis from the study could remove confusion.  The New York Times also omits several molecular mechanisms through which these microdeletions could cause schizophrenia, explained in the Nature article. 

External Sites

The Copy Number Variation Project: http://www.sanger.ac.uk/humgen/cnv/

Schizophrenia Society of Canada: http://www.schizophrenia.ca/heimEnglish1.htm

Works Cited

Eunice Kua, Michael Reder, and Martha J. Grossel. 2004. Science in the News: A Study of Reporting Genomics. Public Understanding of Science. 13: 309–322. Available from: http://bio.davidson.edu/courses/genomics/PopularPress.pdf

Stefansson H, Rejescu D, Cichon S, Pietiläinen OPH, Ingason A, Steinberg S, Fossdal R, Sigurdsson E, Sigmundsson T, Buizer-Voskamp JE. Large recurrent microdeletions associated with schizophrenia. Nature [Internet]. 2008 Sept 11 [cited 2011 Jan 30];(Vol 455).  Available from: http://www.nature.com/nature/journal/v455/n7210/full/nature07229.html
 

Wade N. Gene-Hunters Find Hope and Hurdles in Schizophrenia Studies. New York Times [Internet]. 2008 Jul 25 [cited 2011 Jan 30].  Available from: http://www.nytimes.com/2008/07/31/health/research/31gene.html

Genomics Page
Biology Home Page
beriffe@davidson.edu

Email Questions or Comments.

© Copyright 2011 Department of Biology, Davidson College, Davidson, NC 28035