A number of projects described in the book Bioinformatics have been carried using
the machine learning software environment for biological sequence analysis
developed in collaboration by Net-ID, Inc. and employees
at the Danish Center for Biological Sequence
Analysis.
The foundation for the software environment is based on NetLibs,
an object-oriented library of C++ classes for graphical modeling,
machine learning, and inference developed by Net-ID. The library supports
the hierarchical and recursive implementation of any graphical model
(NNs, HMMs, Bayesian Networks, etc.) together with general
local message-passing algorithms for the propagation of information,
errors, and evidence during inferential/learning processes and dynamic
programming.
Net-Libs provides, among other things, the foundation,
for an HMM simulator and an NN simulator for biological sequence
analysis. The easy-to-use graphical interface for both simulators is in Java.
The software environment runs both under Unix and NT platforms.
In addition, the software environment contains facilities for
manipulating input/output sequences, databases, and files,
as well as libraries of trained models. The libraries include
HMMs for a number of protein families and DNA elements
(promoters, splice sites, exons, etc.) and a number of NNs
for the detection of particular structural or functional signals,
both in protein and DNA sequences.
For more information please contact: admin@netid.com
Previous: Ph.D. Courses over the Internet Next: Bioinformatic related meetings and conferences
|
