Abstract
BackgroundGenomic datasets generated by new technologies are increasingly prevalent in disparate areas of biological research. While many studies have sought to characterize relationships among genomic features, commensurate efforts to characterize relationships among biological samples have been less common. Consequently, the full extent of sample variation in genomic studies is often under-appreciated, complicating downstream analytical tasks such as gene co-expression network analysis.ResultsHere we demonstrate the use of network methods for characterizing sample relationships in microarray data generated from human brain tissue. We describe an approach for identifying outlying samples that does not depend on the choice or use of clustering algorithms. We introduce a battery of measures for quantifying the consistency and integrity of sample relationships, which can be compared across disparate studies, technology platforms, and biological systems. Among these measures, we provide evidence that the correlation between the connectivity and the clustering coefficient (two important network concepts) is a sensitive indicator of homogeneity among biological samples. We also show that this measure, which we refer to as cor(K,C), can distinguish biologically meaningful relationships among subgroups of samples. Specifically, we find that cor(K,C) reveals the profound effect of Huntington’s disease on samples from the caudate nucleus relative to other brain regions. Furthermore, we find that this effect is concentrated in specific modules of genes that are naturally co-expressed in human caudate nucleus, highlighting a new strategy for exploring the effects of disease on sets of genes.ConclusionsThese results underscore the importance of systematically exploring sample relationships in large genomic datasets before seeking to analyze genomic feature activity. We introduce a standardized platform for this purpose using freely available R software that has been designed to enable iterative and interactive exploration of sample networks.
Highlights
Genomic datasets generated by new technologies are increasingly prevalent in disparate areas of biological research
By restricting sample network construction to modules of genes that are naturally co-expressed in human caudate nucleus [21], we find that this degradation is most significant in a neuronal signal transduction module
It is applied to very large datasets, in contrast to clustering procedures that rely upon visual inspection of dendrograms to identify outlying samples. It produces a battery of measures for summarizing the consistency and integrity of genomic datasets, which can be compared across disparate studies, technology platforms, and biological systems
Summary
Genomic datasets generated by new technologies are increasingly prevalent in disparate areas of biological research. Genomic studies capture an enormous amount of information about the molecular organization of biological systems. Understanding this organization poses a challenge for biologists. Biologists’ efforts to explore sample relationships in genomic studies are an integral component of data pre-processing, yet they are often performed in a perfunctory fashion using platform-specific and qualitative criteria. Cluster analysis can be impractical for very large datasets, in which the sheer number of samples obscures the organization and characteristics of a dendrogram and produces ambiguous cluster boundaries
Sign-in/Register to view highlights & summary Sign-in/Register to access full text options 
Free) 
Talk to us
Join us for a 30 min session where you can share your feedback and ask us any queries you have
Schedule a call
