A simple genetic algorithm for biomarker mining

Authors:
Dusan Popovic;Alejandro Sifrim;Georgios A. Pavlopoulos;Yves Moreau;Bart De Moor
Affiliations:
ESAT-SCD / IBBT-KU Leuven Future Health Department, Katholieke Universiteit Leuven, Leuven, Belgium;ESAT-SCD / IBBT-KU Leuven Future Health Department, Katholieke Universiteit Leuven, Leuven, Belgium;ESAT-SCD / IBBT-KU Leuven Future Health Department, Katholieke Universiteit Leuven, Leuven, Belgium;ESAT-SCD / IBBT-KU Leuven Future Health Department, Katholieke Universiteit Leuven, Leuven, Belgium;ESAT-SCD / IBBT-KU Leuven Future Health Department, Katholieke Universiteit Leuven, Leuven, Belgium
Venue:
PRIB'12 Proceedings of the 7th IAPR international conference on Pattern Recognition in Bioinformatics
Year:
2012

Citing 6
Cited 0

Toward memory-based reasoning

Communications of the ACM - Special issue on parallelism
Reducing bias and inefficiency in the selection algorithm

Proceedings of the Second International Conference on Genetic Algorithms on Genetic algorithms and their application
An introduction to genetic algorithms

An introduction to genetic algorithms
Bagging predictors

Machine Learning
Random Forests

Machine Learning
Nearest neighbor pattern classification

IEEE Transactions on Information Theory

Quantified Score

Hi-index	0.00

Visualization

Abstract

We present a method for prognostics biomarker mining based on a genetic algorithm with a novel fitness function and a bagging-like model averaging scheme. We demonstrate it on publicly available data sets of gene expressions in colon cancer tissue specimens and assess the relevance of the discovered biomarkers by means of a qualitative analysis. Furthermore, we test performance of the method on the cancer recurrence prediction task using two independent external validation sets. The obtained results correspond to the top published performances of gene signatures developed specially for the colon cancer case.