Finding a research topic

1. What is a suitable research topic?
2. How to find a topic?
3. Examples of good and bad topics

A. What is a suitable research topic?

1. It should permit discovery of something new
   This isn't a term report. By the end of the semester, you should have found something, through the activities of your own hands, that has never been found before. That's a major goal of this research project
    
2. It should be amenable to discovery through bioinformatics tools
   (Please see the name of the course at the top of this page)
    
3. It should be related to bacteria and/or bacteriophage
   We want to maintain a focus as a class so that we share a common basis for discussion. We have a good common database of bacteria and bacteriophage through Phantome/BioBIKE.
    
4. It should be related in some way to your group topic
   You can learn from your group members and they can learn from you if there is some common core that connects your projects.
    
5. It should be small enough to allow significant progress in the time allowed
   You have only six weeks or so. You can do a lot in six weeks, but in that finite period, you can't reach a satisfying end point if you set the finish line too far away.
    
6. It should be large enough so that your research will be interesting to others
   You will be spending a good deal of time on your topic. Others will spend a bit of time listening to you talk about it. You may presume that everyone likes genome sequences, but go beyond that general common basis and find a topic that will grab you and them.
    
7. Can your report reproduce what you find in some research article?
   No. What was reported in that article is already known, and your calling is to find out something no one else has seen before. On the other hand, it's often very helpful to reproduce the work of others. That way you can test your techniques in a known situation. But don't stop there. Adapt the techniques you find in an article to a different situation. That's the way science works.
    

B. How to find a topic?

1. Extract ideas from the group focus articles
   On the Research Project page, you'll find a link to your group (and a link to group assignments if you need a reminder). On your group page, you'll find multiple articles related in some way to the group topic, often review articles. If you skim these articles, you might find something that sparks your interest. You might also profit from finding other, perhaps better articles yourself. (Here's how to focus a search to review articles).
    
2. Extract references to pertinent articles from review articles and the Introductions to research articles
   One of the primary benefits of a review article is that it provides you with references that someone in the field thinks are significant. Review articles and the reviews found in Introductions can save you hours of searching! Look particularly for research articles that sound like they may approach the problem from a bioinformatic angle (though it's often difficult to tell from just the title).
    
3. Look for a published model of what you want to do
   You'll find nothing more helpful to your project than a good research article. You can find them using methods you already know. Browse many to search for one or more that use bioinformatics tools to address your area of interest.
    
4. Use a result from a research article as a starting point
   • Extend a result: They found that X is true in A... is it also true in B?
   • Refine a result: They found that X is true if you average over all A through F... but maybe X is significantly more true if you average over all A through C and less true averaging over D through F?
   • Analogize a result: They found that X is related to Y... Well, X and Y remind me of P and Q. Maybe P is also related to Q?
      
5. Focus
   It's likely that your first stab at a topic will be too big, something that it would take you months or years to complete. That's OK. But don't let the topic paralyze you! As quickly as possible get to the point where you're doing computation on something specific, looking for a specific sort of sequence on a specific genome. By doing you'll get a sense of what is practical and what is not far faster than you can by reading. Then refine your topic to something more achievable.
    
6. Talk
   Try out your ideas on other people -- me, your group, your mother -- anyone. The more you talk about it, the more it takes a real shape.

C. Examples of bad topics

• "Dispersed repeats in bacterial genomes"
  Way too big. You wouldn't finish this for years. Focusing on a particular type of dispersed repeat would help, and so would focusing on a small group of related bacteria,
   
• "How do phages replicate their DNA?"
  Way too diffuse. Sounds like the topic of a general term report. It's hard to see what you could hope to discover yourself within this huge umbrella.
   
• "The DnaA-binding sites in the origin of replication of Yersinia pestis strain CO90"
  Too small. It would be more interesting to compare the sites from several related strains and/or analyses of sites outside the origin region.
   
• "The big, the bad, and the beautiful"
  Poetic, perhaps, but this topic offers no clue about what you're actually going to study.