• Search |
• Edit this page

Notes on Needleman-Wunsch prototype

• The "next" button really does two things, "go to the next step" and "do it". Maybe there should be two buttons?

Basic programming concepts:

 CONCRETE
    ^
    |
    | conditionals
    |
    | loops
    |
    | functions
    |
    V
 ABSTRACT

Selling Points:

Needleman-Wunsch does not deal well with re-arrangements. I think BLAST might be better, but biologists don't know this sort of thing.

However, when they see N-W... within minutes they ask the question. They see how simple the algorithms are and they lose their respect for them.

In a sense, they thing bioinformaticists are gods. They think the tools are magic. This is dangerous at best, and counterproductive at worst.

• Question: Does ClustalW use BLAST or Smith-Waterman? Or something else?

Critiques:

Yes, you could provide a pamphlet. Yes you can try to improve education. But you are only explaining the factors that YOU, the bioinformaticist have deemed are relevant. The trouble is, you don't know what's relevant to every biologist. They have unique needs (do they?). This is especially true when it comes to something as prolific as sequence alignment.

I think the likes of Mehmet REALLY want to see a good comparison to the pamphlet. I might be able to get away with not doing a quantitative experiment, but I definitely need some good evidence.

Alternatives:

• progressive disclosure of code (collapsing code tree)
• practical glossary. i.e., you look up a term and it tells you what you need to know about it.
• pipeline constructor (like stuff mentioned in Phone call with David Wild 11-11-05) w/ good explanations
• simple app that does one thing w/ good design

Ideas

• search for terms, code, etc

Other stuff

PIN-IT makes inferences, and then it shows how they were calculated. This is exposing the biologists to technical details...

http://homer.informatics.indiana.edu/~raod/query.cgi