199

capable partners with computers, and should actually use their human foresight to verify

the computer’s calculated results, rather than blindly (or even “subordinately”) believing

them. Needless to say, in fact, the relationship between the two subjects is rather cordial.

Artificial intelligence research and bioinformatics intensively cross-fertilize each other.

For example, a number of sophisticated search strategies of artificial intelligence have

been inspired by biological phenomena. Conversely, neuronal networks (and also Hidden

Markov Models; Chap. 3) are used for many sequence or even more complex predictions

in bioinformatics (e.g. for signal sequences).

14.1

14.1a

­

14.1b

Some Cornerstones of Artificial Intelligence

(a) Principle: neural networks in general

(b) Software that uses neural networks:

NucLocP; SignalP or for transmembrane proteins: TMHMM

14.1

(d) Expert systems

–– Example area medical informatics:

e.g. infections, burns, anaesthesia

–– Wolfram Mathematica

14.1  People Question, Computers Follow Programs