Course projects for CPSC 545:

Every student needs to complete a course project for CPSC 545. The course project should either be review paper (option 1) or a programming project (option 2).

Option 1: Review paper

  • purpose: survey an interesting area of ongoing research, take a position on it and propose and discuss potential future research directions
  • group: teams of two students
  • proposal: 1-2 pages of text (excluding references) plus at least five key references
  • interim report: email me a brief status report (1 page). This report should mention (1) what you have achieved so far, (2) if there are any problems and (3) what your plan for the rest of the time is. Please attach to your email the version of the document as it is at that point.
  • final paper: review paper of 8--10 pages (excluding references) in the style of a research paper
  • problems: if you think there is a problem with your project that requires my input, please get in touch with me as early as possible (e.g. email)

Option 2: Programming project

  • purpose: explore an interesting and novel idea, implement it into a new or existing piece of Bioinformatics software and validate its performance
  • group: teams of two students
  • proposal: 1-2 pages of text (excluding references) plus at least five key references
  • interim report: email me a brief status report (1 page). This report should mention (1) what you have achieved so far, (2) if there are any problems and (3) what your plan for the rest of the time is.
  • final implementation and report: well-documented Java or C++ source code (including readme and make files for a Linux machine, scripts may be written in Perl or Python) and a project report of 8--10 pages (excluding references) in the style of a research paper

Writing the project proposal:

Your project proposal should be sent as pdf file and should:

  • have a meaningful title,
  • list your group members,
  • motivate your project (1/2 page),
    • Why is this interesting (for you and the research community) ?
    • What are you hoping to learn and achieve ?
  • summarize the state of art in the field (1/2 to 1 page),
    • What has been done already (identify and cite at least 5 key references) ?
    • What are ongoing research directions ? In how far is this an active area of research ?
    • What are people in the field currently trying to achieve ?
  • outline of what your review is intended to cover and achieve (option 1: review paper) or list the milestones of your project with a rough time line (option 2: programming project) (1/2 to 1 page), in particular
    • Option 1: Which ideas for potential future research directions (algorithms, computational methods, data analyses etc) do you hope to propose ?
    • Option 2: Justify that your project does not contain strategic bottlenecks which are likely to render the entire project infeasible (e.g. availability of data sets or CPU power for analysis) and take care to clearly identify a few, well-defined and realistic research goals.

Writing the final project report:

The final report of your course project should be in the style of a research paper. This tar-ball contains a latex outline of a research paper for the Bioinformatics journal which may help you set up your own document. Feel free to adjust the proposed structure of the paper to your own type of research, for example by moving the section on methods from the end to the middle of the paper if you propose a new algorithm or computational method.

If your course project is a review paper, structure your document like a review paper in Bioinformatics rather than a research paper. Make sure you devote no more than 2/3 of the pages on the review of existing methods (including a detailed and self-consistent description of all computational techniques) and at least 1/3 of the pages on your ideas for novel research directions (algorithms, data analyses, new computational methods). This guidelines apply to both type of course projects.

Selecting a course project:

Both types of course projects provide you with an opportunity to explore one area of Bioinformatics research in more detail.

Option 1, the review paper, is meant to let you explore one ongoing area of Bioinformatics research in greater depth by

  • reviewing the relevant literature,
  • summarizing and judging the research conducted so far,
  • explaining the different computational techniques and algorithms involved in detail and
  • proposing and discussing interesting ideas for future research directions.

Each review paper is meant to be written by a team of two students who will receive the same grade for their project. Please note that each topic can only be covered by at most one team.

Option 2, the programming project, is an opportunity to explore a novel idea for a small, well-defined research project in Bioinformatics in practice by

  • implementing a new idea into new or existing source code (or, alternatively, proposing new algorithms and comprehensively exploring their theoretical properties only mathematically),
  • evaluating the performance of the new approach and
  • presenting the entire project in a comprehensive and self-contained project report.

The programming project should be done by teams two students who will receive the same grade for their project. When writing your project proposal, make sure that you outline a realistic project plan that can be implemented, evaluated and documented in the proposed time frame. In particular, take care to identify and eliminate potential bottlenecks (e.g. availability of data) which may render your entire project infeasible (and very frustrating).

Potential topics for review papers which would fit the overall scheme of this course are, to name just a few in no particular order:

  • efficient parameter training algorithms for hidden Markov models
  • efficient parameter training algorithms for stochastic context free grammars
  • using high-throughput sequencing data for improving gene prediction methods
  • computationally efficient methods for RNA structure prediction including pseudo-knots
  • alignment free methods for comparative RNA structure prediction
  • hidden Markov models versus conditional random fields for gene prediction
  • predicting RNA editing sites (with and without predicting or taking known local RNA structure into account)
  • promoter prediction (comparative and non-comparative approaches)
  • ...

The following journals or conference proceedings may provide you with inspiration for course projects:

The following resources are helpful for literature searches:

  • Pubmed: use, for example, "Keyword 1"[title] AND "Keyword 2"[title] for searching which will also retrieve and highlight review papers
  • ISI Web of Knowledge: this commercial web site allows for the most detailed literature searches and is also capable of retrieving recent papers which cite a given, older paper; make sure you connect via a UBC machine when using this data base
Updated: December 16, 2011, Irmtraud Meyer