Technical Report Details
||A Regularized Method for Selecting Nested Groups of Relevant Genes from Microarray Data
||Christine De Mol, Sofia Mosci, Magali Traskine, Alessandro Verri
author = "De Mol, C. and Mosci, S. and Traskine, M. and Verri, A.",
||Gene expression analysis aims at identifying the genes able to
accurately predict some biological parameter like, for example,
disease subtyping or progression. While accurate prediction
can often be achieved by means of many techniques, gene
identification, due to correlation and the limited number of
available samples, is a much more elusive problem.
Small changes in the expression values often produce different
gene lists and solutions which are both sparse and stable are difficult to
obtain. We propose a consistent two-stage regularization
method able to learn linear models characterized by high prediction
performance. By varying a regularization parameter the linear models
trade sparsity for the inclusion of correlated genes and produce gene
lists which are almost perfectly nested.
Experimental results on benchmark microarray confirm the interesting
properties of the proposed method and its potential as a starting
point for further biological investigations.