Loading...
Please wait, while we are loading the content...
The Use of a Self Organizing Map to Generate QSAR Sets
| Content Provider | Semantic Scholar |
|---|---|
| Author | Guha, Rajarshi |
| Copyright Year | 2003 |
| Abstract | I have been working on two aspects of QSAR modelling model generation and development of algorithms. In the former area I have been involved in a study of artemisinin analogues. Artemisinin is a well known anti-malarial. However certain analogues are neurotoxic and much research has been carried out to synthesize non neurotoxic analogues. The compounds in my current study have been studied using the CoMFA methodology. My goal was to investigate whether the analogues could be modelled using the simpler ADAPT methodology. This is ongoing work and current results seem to indicate that the computational neural networks are able to produce good predictive models. In this report I will be focussing on my work with the Kohonen self organizing map (SOM) and its use in the generation of representative training, cross-validation and prediction sets (collectively referred to as QSAR sets). |
| File Format | PDF HTM / HTML |
| Alternate Webpage(s) | http://cheminfo.informatics.indiana.edu/~rguha/writing/pub/somqsar-rep.pdf |
| Alternate Webpage(s) | http://www.rguha.net/writing/pub/somqsar-rep.pdf |
| Language | English |
| Access Restriction | Open |
| Content Type | Text |
| Resource Type | Article |
