KEDRI - Gene Expression Profiling Siftware

KEDRI
AUT

KEDRI
AUT Main Nav
Centre Banner
Main Content

Gene Expression Profiling Siftware

Gene Expression Profiling Siftware

Siftware Team

Introduction

Gene expression data is frequently of a high dimensional nature. Coupled with a small number of samples, this can make it hard to identify which genes are important for the development of a disease classification model. Here, a software package is presented which is designed to help the user to visualise, analyse and model this data.

Siftware Main Interface.
fig 1. Siftware Main Interface

Gene Expression Profiling Siftware

This software breaks the problem down into 3 stages: Visualisation. This section includes 3D visualisation and Principal Component Analysis Data analysis. This section includes several dimension reduction mechanisms, Signal to Noise Ratio (SNR) and correlation coefficient analysis. It also contains several clustering algorithms, K-means clustering and Hierarchical clustering.

Signal to Noise Ratio.
fig2. Signal to Noise Ratio

Modeling and discovery. This section includes cross validation for modelling and feature selection in an unbiased way including the essential cross validation module from NeuCom project. The available methods are: Signal To Noise Ratio, T-Test, Multiple Linear Regression, Support Vector Machine, K Nearest Neighbour, Weighted K Nearest Neighbour, Multi-Layer Perceptron, Radial Basis Function, Evolving Classification Function and Evolving Clustering Method for Classification.

Best combination of genes.
fig3. Best combination of genes

Conclusions

The software presented here represents a novel and comprehensive approach to modeling and visualising gene expression data. It also provides a concise method of viewing results. Further work in this environment will include the application of Evolutionary Computation for the optimisation of ECOS modeling procedures.

References

[1] 2001 - Kasabov, N., Evolving Fuzzy Neural Networks for Supervised/Unsupervised On-Line, Knowledge-Based Learning, IEEE Transactions on Systems, Man and Cybernetics, Part B: Cybernetics, Vol 31, No. 6 Issue (December 2001, pp.902-918)

[2] 2001-Kasabov, N., Adaptive learning system and method, WO 01/78003, under PCT (publication date 20.04.2001)

[3] 2002-N.Kasabov, M. Futschik, M.Sullivan, A.Reeve, A method and system for using microarray gene expression data and clinical information, Preliminary application, 2002, USA

[4] 2002- N.Kasabov, A.Reeve, M. Futschik, M.Sullivan, P. Guildford, Medical Applications of Adaptive Learning Systems, Preliminary Patent Application, February 2002, New Zealand


Last updated: 02 Dec 2011 10:07am

Auckland University of Technology, New Zealand | Copyright © | Privacy | Site map