Accurate modelling of biological systems requires a deeper and more complete knowledge about the molecular components and their functional associations than we currently have. Traditionally, new knowledge on protein associations generated by experiments has played a central role in systems modelling, in contrast to generally less trusted bio-computational predictions. However, we will not achieve realistic modelling of complex molecular systems if the current experimental designs lead to biased screenings of real protein networks and leave large, functionally important areas poorly characterised. To assess the likelihood of this, we have built comprehensive network models of the yeast and human proteomes by using a meta-statistical integration of diverse computationally predicted protein association datasets. We have compared these predicted networks against combined experimental datasets from seven biological resources at different level of statistical significance. These eukaryotic predicted networks resemble all the topological and noise features of the experimentally inferred networks in both species, and we also show that this observation is not due to random behaviour. In addition, the topology of the predicted networks contains information on true protein associations, beyond the constitutive first order binary predictions. We also observe that most of the reliable predicted protein associations are experimentally uncharacterised in our models, constituting the hidden or "dark matter" of networks by analogy to astronomical systems. Some of this dark matter shows enrichment of particular functions and contains key functional elements of protein networks, such as hubs associated with important functional areas like the regulation of Ras protein signal transduction in human cells. Thus, characterising this large and functionally important dark matter, elusive to established experimental designs, may be crucial for modelling biological systems. In any case, these predictions provide a valuable guide to these experimentally elusive regions.

MOTIVATION: A recent development in sequence-based remote homologue detection is the introduction of profile-profile comparison methods. These are more powerful than previous technologies and can detect potentially homologous relationships missed by structural classifications such as CATH and SCOP. As structural classifications traditionally act as the gold standard of homology this poses a challenge in benchmarking them. RESULTS: We present a novel approach which allows an accurate benchmark of these methods against the CATH structural classification. We then apply this approach to assess the accuracy of a range of publicly available methods for remote homology detection including several profile-profile methods (COMPASS, HHSearch, PRC) from two perspectives. First, in distinguishing homologous domains from non-homologues and second, in annotating proteomes with structural domain families. PRC is shown to be the best method for distinguishing homologues. We show that SAM is the best practical method for annotating genomes, whilst using COMPASS for the most remote homologues would increase coverage. Finally, we introduce a simple approach to increase the sensitivity of remote homologue detection by up to 10%. This is achieved by combining multiple methods with a jury vote. SUPPLEMENTARY INFORMATION: Supplementary data are available at Bioinformatics online.