Biological sequence analysis and information processing by artificial neural networks Søren Brunak
Center for Biological Sequence Analysis
Technical University of Denmark
brunak@cbs.dtu.dk
Biological sequence analysis and information processing by artificial neural networks Søren Brunak
Center for Biological Sequence Analysis
Technical University of Denmark
brunak@cbs.dtu.dk
Biological neuron
Diversity of interactions in a network enables complex calculations
Similar in biological and artificial systems
Excitatory (+) and inhibitory (-) relations
between compute units
Transfer of biological principles to neural network algorithms
Non-linear relation between input and output
Massively parallel information processing
Data-driven construction of algorithms
Ability to generalize to new data items
Simplest non-trivial classification problem
CNHSYYP, HIETRRA, NWQSADY, NQYSEPR, WHITRCA, DYHSANY, ...
Two categories: positives and negatives
Data described by two features, e.g.
charge, sidechain volume, molecular
weight, number of atoms, ...
Comments