Journal Club

One goal for our introductory genetics course is to introduce students to reading the primary research literature.  Many students find journal articles to be daunting, thinking that the information in a scientific paper is too complex for them to understand. The directed journal club is one strategy that we use to help the students become more capable of, and comfortable with, reading the primary literature.  The journal club feature available on-line offers some guidance for instructors who would like to implement directed journal clubs in their own classes. The research papers used for the journal club feature are related to concepts introduced in the book, and concern topics that many students find engaging.  We have also endeavoured to select articles that do not require too much specialist knowledge in disciplines beyond the scope of the text.  We find that students gain a real feeling of achievement when they realize that they have the tools and ability to understand and even critique a scientific paper.

In a directed journal club in our course, we provide the students a link to the journal article and a set of study guide questions several days before class. We then write the numbers of the questions on index cards and distribute the cards to students at the class time.  Students are given a few minutes to compose their answers, and are allowed to exchange questions with another student if both agree. We expect the students to be able to provide an oral answer to the study guide question they receive, often with a very brief explanation of less than five minutes.  Although we expect the students to be prepared to provide an answer to all or nearly all of the study guide questions, no student is asked to explain the entire paper, but every student is expected to be able to explain part of it.  Students are encouraged to ask questions of their peers and add comments of their own to the class discussion.  All of the figures and tables from the paper are available in class for the students to use as visual aids during their oral presentations.

In preparing the questions for the Journal Clubs, we have tried to anticipate some of the challenges that face students as they learn to read the primary literature.  Some questions can be readily answered with facts in the paper and require little additional research or critical thinking.  Some ask for the explanation of a figure, as we find that students need practice in the careful reading of graphs and tables.  Some questions focus on techniques, while other questions encourage the students to think about the bigger intellectual questions. A few require the student to look up a previous paper to learn how the experiment was done.  The single biggest challenge for students is to distinguish the actual experimental results from the authors’ inferred conclusions.  We have found that questions requiring the students to explain the control experiments are very helpful in this regard.

Therefore, in addition to the prepared questions we find it helpful to encourage class discussion after each student answer. We carefully monitor student answers to make sure they: (i) really have identified the question being addressed in a particular experiment, (ii) that they understand what the data actually shows as a separate step from the interpretation and conclusions of an experiment (iii) they understand the main conclusion being drawn and (iv) that they have identified controls and understand why they have been included. Follow-up questions that further clarify these points often enhance the discussion. We also encourage students to tell us what they found difficult to follow or understand. Students sometimes fixate on small details of lesser importance that they don’t understand at the expense of the main points. Guiding the discussion in a way that models how we ourselves navigate complicated experimental data is often very helpful.

The journal club feature provides examples for instructors; we hope you will use these as a starting point and adapt this approach to your particular course.  There are and will be many other papers that can be used to stimulate student interest and engagement in the primary research literature.

Chapter 10 Journal Club

Paper:  Sutter et al., 2007.  A single IGF1 allele is a major determinant of small size in dogs.  Science 316: 112-115.

Background information:

This paper uses the principles of positional cloning and genome association studies to identify a gene that appears to play a significant role in the size of different dog breeds.  In addition to working with a familiar and easily measured phenotype, the paper illustrates many of the important techniques and strategies that are important in genome-wide association studies (GWAS) and other mapping and association-based approaches to find the causative genes for phenotypes.  Once you have read the paper and supporting information, be prepared to answer the following questions.  Solutions to the questions are available to registered instructors.

We suggest the following strategy for reading this and many scientific papers.

  • First, read the paper for its general conclusions.  You may need more than one reading for this.  Take notes. This probably will not take very long—more than 15 minutes but probably less than 30 minutes for most papers. You are trying to get the big picture rather than all of the details.
  • Second, examine each of the figures closely and try to understand what each one is showing. Try to determine what questions was being asked by the experiments, how the experiment was done, and what controls were included.  Try to draw your own inferences from the data before looking at the authors’ conclusions.  Examine the figure legends for help.  Look up terms in the figures that are not familiar to you. This will probably take the most time and effort.
  • Third, attempt to answer each of the questions as best you can, and keep notes on your answers.  In our experience, students often like to work together on this so long as each person is prepared to answer the questions.   All researchers routinely discuss papers with one another to learn from each other’s insights.  We strongly encourage our students to work cooperatively on these questions.

Questions:

  1. Other than a general interest in dogs, what are some of the important biological concepts that might be learned from studying size diversification in dogs?
  2. Prior to this study, what other hypotheses had been suggested to explain the size diversification in dogs, and how might each of these ideas account for size differences?  This will require you to look up and briefly examine some references in addition to this paper.
  3. Why were the initial studies based on Portuguese water dogs and what would be the advantage of doing the initial studies within a single breed
  4. Is this study a true GWAS?  If this paper is not a “genome-wide association”, what led them to focus on this region of the genome?
  5. Mapping methods to identify candidate genes for a phenotype depend on having appropriate and easily scored markers nearby.  Concisely but completely explain how they identified such markers for these experiments.  Could these same markers be used for mapping and association studies for other traits in other breeds?  Why or why not?
  6. Explain the experiment and the results shown in Figure 1A.  What is the conclusion that can be drawn from this graph?
  7. Explain the experiment shown in Figure 1B and discuss what information this adds to the results shown in Figure 1A.  (Note from the erratum with the paper that two of the labels were reversed so be sure to refer to the correct labels.)
  8. Figure 1C shows the serum levels of IGF for each of the three genotypes.  What additional information does this experiment add to what is shown in Figures 1A and 1B?  Assuming that IGF1 levels are causative for this trait, does the difference appear to behave as a dominant, a co-dominant, an intermediate dominant, or a recessive allele?
  9. When they turned from analyzing only Portuguese Water Dogs to including other breeds as well, two important and related concepts were introduced.  Explain the terms “linkage disequilibrium” (abbreviated LD) and “signature of selection”, and discuss what role these have played in their experiments.
  10. When we think about all breeds of dogs (particularly including non-purebreds), size variation probably follows a continuous distribution.  How did they classify small and large breeds?  What are some advantages to the researchers to using an experimental strategy that places phenotypes into two non-overlapping categories like this?  We return to this concept later.
  11. Two technical and possibly unfamiliar terms are introduced in Figure 2.  “Heterozygosity” is a measure of how many individuals in a population are heterozygous for alleles or polymorphisms at a particular locus.  “FST” (the fixation index) is a measure of how much of the variation in a population can be explained by some feature of the population structure, such as individuals being related to one another or descended from a common ancestor.  (An FST of 0 implies that no individuals in the population are related to each other, while an FST of 1 indicates that every individual in the population is genetically related to every other individual.)  With that in mind, briefly summarize the results shown in Figure 2 and the main conclusions that can be drawn from these results.
  12. Figures 1A and 2A, B, C, and D are graphing the associations of polymorphisms at various positions on chromosome 15 with variations in the phenotype.  However, the scale on the X axis indicates that Figure 1A shows a much larger region of the chromosome than do the panels of Figure 2.  What is the explanation for this change in the scale of the X axis?
  13. What is the main conclusion that can be drawn from the results shown in Figure 3A, and what controls are being introduced in this figure? The statistical analysis is complex, so unless your instructor tells you otherwise, a descriptive and non-quantitative answer is acceptable.
  14. Figures 3B and 3C form one large panel that reads from left to right.  In other words, Haplotype A in Figure 3B accounts for the top line of the table in Figure 3C.  Explain these results carefully and fully.  In our experience, many students understand the concept of a haplotype much more fully after analyzing this figure.
  15. Large and small breeds of dogs do not show the same strength of association with a particular haplotype.  That is, whereas nearly all small breeds are associated with Haplotype B, many but not all large breeds are associated with Haplotype I.  What does this result suggest about the underlying biological process that led to the differentiation between small and large breeds?  What other evidence in this paper is also consistent with these results, and might have led to the hypothesis that large breeds are somewhat more variable than small breeds?
  16. Briefly discuss the cellular and physiological function of IGF1 and some of the evidence that it is involved in size differentiation in mammals. This may require you to use additional references.
  17. Although size diversification is strongly associated with the IGF1 gene, did they in fact identify a causative mutation?  What additional experiments might need to be done to demonstrate conclusively that IGF1 changes are causative for the size diversification?
  18. Which of sequence differences identified in the IGF1 gene is likely to be the causative mutation, in your estimation?  Is there any evidence in the paper that supports your hypothesis?
  19. Based on this study, briefly explain the origin of small dog breeds, using the supporting evidence from this paper.
  20. In question 10, we noted that the investigators defined two non-overlapping phenotypes for dog size, and that this provided certain advantages for their study.  Geneticists often use a binary description of phenotypes like this in segregation analysis of traits; for example, Mendel referred to “tall” and dwarf” plants and “round” and “wrinkled” seeds without considering any intermediate phenotypes.  Imagine that these or other investigators had included intermediate size phenotypes on many different dog breeds (or mixed breeds) and had done a genome-wide association.  Which of these statements comes closest to the expected results from such a study? Note that most of these hypotheses are plausible so this question attempts to tune your biological judgment about the relative impact of each effect.  Explain your reasoning.
    • Nearly all of the variation can be explained by molecular changes in the IGF1 gene but there will be different molecular changes. Among different dogs, there will be splice variants, amino acid substitutions, regulatory changes, and so on but nearly all of these will be in IGF1.
    • A sizeable fraction of the variation in phenotype can be explained by molecular variation in IGF1. Nearly all of the rest of the phenotypic variation is associated with molecular differences in other genes in the insulin signaling pathway.
    • Changes in IGF1 and other genes in the insulin signaling pathway will account for a substantial amount of the phenotypic variation, but an unknown number of other genes will also be found to affect the phenotype.
    • IGF1 and other genes in the insulin signaling pathway will account for only a very small amount of the variation in size, and genes not related to IGF1 will account for most of the variation.
    • Size variation will depend on the breed. Among these breeds, variation in IGF1 is significant, but among some others breeds, other genes will be more significant.