Applications
of
Constraints
are
Everywhere

[Current Week]

Week of November 4, 1996

Web page maintenance

• Added Stephen to home page
• Created 'weekly log' template files

ISMB-95 Abstracts review

• Reviewed the abstracts from the Proceedings of the Third International Conference on Intelligent Systems for Molecular Biology to determine if it would be worthwhile to order the proceedings, and to become familiar with some of the latest research directions and groups.
  • APIC : A Generic Interface for Sequencing Projects
    Describes a standard graphical interface for displaying the results of varied genomic sequence analysis methods and thus making comparisons easier. Not useful to us directly, but perhaps to display and compare the results of any such analyses we might experiment with.

  • A Specification for Defining and Annotating Regions of Macromolecular Structures
    Program- and machine-independent standard for annotating macromolecular structures; easy to parse, but also human-readable. This might be quite interesting, especially for what I was thinking with using an LP system to encode higher-level 'empirical' or 'expert' knowledge; this could be the standard language that researchers could use to encode their expert knowledge.

  • Predicting Protein Folding Classes without Overly Relying on Homology
    A "novel" approach to predicting structural folding classes from primary sequences. Might be interesting, but not enough info in abstract to tell.

  • Neural Net Representations of Empirical Protein Potentials
    Using neural nets to apply "knowledge-based" empirical potential functions for proteins. Appears to be a higher-level, more generalized approach than other current ones. Might be interesting, just to see what some 'competing' groups are doing in this area, and to get some ideas.

  • Classifying Nucleic Acid Sub-Sequences as Introns or Exons Using Genetic Programming
    Little info in abstract, but I would like to read about some more practical applications of genetic programming techniques.

  • Constituting a Receptor-Ligand Information Base from Quality-Enriched Data
    Like the "Specification for Defining and Annotating..." (above) it might be interesting to watch if this becomes adopted as a standard database format for information required for drug design.

Week of November 11, 1996

Topic

• Reviewed ISMB-95 Tutorial T5 - The Computational Linguistics of Biological Sequences (David Searls)

Week of November 18, 1996

Topic

• Reviewed ISMB-96 Tutorial T2 - Machine Learning for Molecular Biology (Steven Salzberg and Jude Shavlik)

  The tutorial slides, even with the included speaker notes, were too lacking in information to be very useful, or even understandable. I think they would only be of value to someone who had actually heard the presentations. This is somewhat different from some of the other tutorials from this conference. However, an appendix to this tutorial includes an 'annotated bibliography' and a survey of web-based resources that might be a very good starting point for further literature reviews.

Week of November 25, 1996

Topic

• Reviewed Integrative Approaches to Molecular Biology (Julio Coolando-Vides, Boris Magasanik, Temple F. Smith, eds.)

  Generally, appears to be a very interesting book. However, it mainly focuses on high-level, sometimes almost philosophical, integrative approaches, and at this time I'm looking for some more 'hard details' about what various research groups are doing. This is not to say the book is not useful; in fact, it presents some very interesting ideas, and might be a good 'launching pad' for new research directions.

  The book is a collection of papers, and two of the chapters, 3 and 15, provide a good survey of some of the current techniques and thoughts on protein structure predection and grammatical analysis of DNA sequences, respectively. Anyone interested in these topics should find these two to be good starting points for literature reviews.