Jove
Visualize
Contact Us
JoVE
x logofacebook logolinkedin logoyoutube logo
ABOUT JoVE
OverviewLeadershipBlogJoVE Help Center
AUTHORS
Publishing ProcessEditorial BoardScope & PoliciesPeer ReviewFAQSubmit
LIBRARIANS
TestimonialsSubscriptionsAccessResourcesLibrary Advisory BoardFAQ
RESEARCH
JoVE JournalMethods CollectionsJoVE Encyclopedia of ExperimentsArchive
EDUCATION
JoVE CoreJoVE BusinessJoVE Science EducationJoVE Lab ManualFaculty Resource CenterFaculty Site
Terms & Conditions of Use
Privacy Policy
Policies

Related Experiment Videos

Multiclass cancer classification using gene expression profiling and probabilistic neural networks.

Daniel P Berrar1, C Stephen Downes, Werner Dubitzky

  • 1School of Biomedical Sciences, University of Ulster at Coleraine, BT521SA, Northern Ireland. dp.berrar@ulster.ac.uk

Pacific Symposium on Biocomputing. Pacific Symposium on Biocomputing
|February 27, 2003
PubMed
Summary
This summary is machine-generated.

Related Concept Videos

You might also read

Related Articles

Articles linked to this work by shared authors, journal, and citation graph.

Sort by
Same author

Deep learning in bioinformatics and biomedicine.

Briefings in bioinformatics·2021
Same author

Computational Systems Biomedicine.

Briefings in bioinformatics·2016
Same author

Folic Acid Supplementation in Postpolypectomy Patients in a Randomized Controlled Trial Increases Tissue Folate Concentrations and Reduces Aberrant DNA Biomarkers in Colonic Tissues Adjacent to the Former Polyp Site.

The Journal of nutrition·2016
Same author

Anatomy and Physiology of Multiscale Modeling and Simulation in Systems Medicine.

Methods in molecular biology (Clifton, N.J.)·2015
Same author

Evaluating a common semi-mechanistic mathematical model of gene-regulatory networks.

BMC systems biology·2015
Same author

From single cells to tissues: interactions between the matrix and human breast cells in real time.

PloS one·2014

A novel probabilistic neural network (PNN) model effectively classifies multiclass cancer gene expression data. This PNN model addresses high dimensionality, noise, and misclassification costs, outperforming traditional methods.

Area of Science:

  • Bioinformatics
  • Computational Biology
  • Machine Learning

Background:

  • Gene expression profiling using microarrays is crucial for cancer classification and diagnosis.
  • Existing machine learning methods struggle with high-dimensional, noisy microarray data and lack robust confidence measures or cost-sensitive learning.

Purpose of the Study:

  • To introduce a probabilistic neural network (PNN) model designed to overcome limitations of current methods in analyzing gene expression data.
  • To evaluate the PNN model's performance in multiclass cancer classification, considering statistical confidence and misclassification costs.

Main Methods:

  • Development and application of a probabilistic neural network (PNN) model tailored for high-dimensional, noisy gene expression data.
  • Comparison of the PNN model against a decision tree and a standard neural network using a lift-based scoring system for performance evaluation.

Related Experiment Videos

Main Results:

  • The PNN model demonstrated superior performance in multiclass cancer classification compared to decision tree and neural network methods.
  • The PNN model successfully addressed challenges of high dimensionality, noise, and incorporated asymmetrical misclassification costs.

Conclusions:

  • The probabilistic neural network (PNN) model is a powerful and statistically sound tool for multiclass cancer classification using gene expression data.
  • The PNN model offers significant advantages over existing machine learning techniques for complex biological data analysis.