Applying text mining to search for protein patterns

Abstract

In this work the problem associated to the obtaining of protein patterns associated with certain cancer types starting from biomedical texts is presented. The research is based on the study of the application of text mining and retrieval techniques to biomedical texts and its adaptation to this problem.Our goal is to annotate a significant corpus of biomedical texts, select the more relevant ones and to train machine learning methods to automatically categorize them along certain dimensions that we have previously defined. The idea behind this project is to identify a group of proteins associated with different cancer types. © 2009 Springer Berlin Heidelberg.