A new statistical method that identifies similar groups of genes from large tissue samples could be used by health care professionals to predict the effectiveness of treating patients suffering from a variety of disorders including cancer, according to a research article published in Genome Biology today. The statistical method, named gene-shaving, because it cuts away useless or irrelevant data to leave clusters of similar genes, was used by the research team at Stanford University, USA, to analyse the genetic information of tumors removed from patients suffering from diffuse large cell B-lymphoma (DLCL). The researchers found that they could identify distinct clusters of genes that, from previous studies, were highly indicative of a patient's chance of survival.
"One important motivation for developing gene shaving was the wish to identify distinct sets of genes whose variation in expression could be related to a biological property of the samples," says Robert Tibshirani, one of the lead researchers from the Stanford University project team. "In the present example, finding genes whose expression correlates with patient survival is an obvious challenge."
The team used more than 4,500 gene expression measurements taken from tumours of 48 patients with DLCL. The measurements were gathered using new DNA array techniques that allow researchers to gather detailed genetic information from a tissue sample. "A major challenge in interpreting these results is to understand the structure of data produced by such studies, which often consist of millions of measurements," continues Tibshirani.
The gene shaving process involves repeatedly applying a complex mathematical formula to the large sea of genetic expression data in order to find the clusters, or islands, of genes that are useful to the researchers. By selecting the clusters of genes that include the strongest individual genes for predicting survival, researchers have been able to identi
Contact: Andrew McLaughlin