399.txt

This complete matlab for neural network

发信人: GzLi (笑梨), 信区: DataMining
标  题: Automated drug discovery via datamining
发信站: 南京大学小百合站 (Tue Dec 17 22:10:23 2002)


A project on automated drug discovery via data mining . 
http://www.rpi.edu/locker/82/001182/public_html/files/

The Automated Design and Discovery of Novel Pharmaceuticals 
using Semi-Supervised Learning in Large Molecular Databases
This interdisciplinary research project is jointly funded in three NSF direct
orates: CISE/IIS, BIO/DBI and ENG/BES. The techniques developed in this resea
rch result in a new framework for the virtual discovery of new pharmaceutical
s or materials. The 
basic idea is to utilize large existing pharmaceutical databases as input for
 a new type of structure/activity correlation methodology in order to calcula
te a large set of new and traditional descriptors to create improved Quantita
tive 
Structure-Activity Relationship (QSAR) models that characterize and predict i
mportant biological responses. 
      Once the descriptors have been determined and a predictive model has be
en built, thousands of new potential molecules, chemically similar to those o
f the benchmark data set, are scanned from large databases and are evaluated 
for their chemical 
properties based on the predictive model. The aim is to target a few novel mo
lecules with potentially attractive pharmaceutical properties that can then b
e tested further in the traditional way in the laboratory. Computationally in
telligent data mining 
techniques are vital to extract the information necessary to select these nov
el molecules. This research applied novel machine learning paradigms such as 
semi-supervised learning with capacity control. These algorithms predict desi
red biological 
responses and generate QSAR models using both known (labeled) and unknown (un
labeled) biological responses. This project involves the development of an in
frastructure of computationally intelligent computer codes that allow for the
 virtual design of 
novel pharmaceuticals or the improvement of existing pharmaceuticals. The pro
posed methodology is applicable to most pharmaceuticals for which a database 
of responses is available. The ultimate pay-off of this methodology is expect
ed to lead to the 
rapid invention of new drugs for new or known society threatening diseases wh
ere a very fast response is warranted.

 --               ***  端庄厚重 谦卑含容 事有归着 心存济物  *** 数据挖掘  htt
p://[e-mail]/ 

※ 来源:．南京大学小百合站 bbs.nju.edu.cn．[FROM: 211.80.38.17]