Quality analysis and integration of large-scale molecular data sets

Abstract

One of the major challenges in bioinformatics today is to integrate and interpret the heterogeneous biological data that are being produced at an ever increasing pace. As this type of analysis is still in its infancy, all studies so far have relied on applying simple rule-based criteria on only a small subset of the available data. To enable comprehensive studies to be undertaken with a statistical framework, standardized repositories from which all datasets can be easily obtained and benchmarks that quantify the often high error rates of large-scale datasets are needed. Quality control, benchmark and integration efforts from protein interaction networks in the context of genome and transcriptome data are reviewed.