Continuing improvements in DNA sequencing technologies are providing us with vast amounts of genomic data from an ever-widening range of organisms. The resulting challenge for bioinformatics is to interpret this deluge of data and place it back into its biological context. Biological networks provide a conceptual framework with which we can describe part of this context, namely the different interactions that occur between the molecular components of a cell. Here, we review the computational methods available to predict biological networks from genomic sequence data and discuss how they relate to high-throughput experimental methods.

The rapidly increasing amount of publicly available knowledge in biology and chemistry enables scientists to revisit many open problems by the systematic integration and analysis of heterogeneous novel data. The integration of relevant data does not only allow analyses at the network level, but also provides a more global view on drug-target relations. Here we review recent attempts to apply large-scale computational analyses to predict novel interactions of drugs and targets from molecular and cellular features. In this context, we quantify the family-dependent probability of two proteins to bind the same ligand as function of their sequence similarity. We finally discuss how phenotypic data could help to expand our understanding of the complex mechanisms of drug action.

To understand a biological process it is clear that a single approach will not be sufficient, just like a single measurement on a protein--such as its expression level--does not describe protein function. Using reference sets of proteins as benchmarks different approaches can be scaled and integrated. Here, we demonstrate the power of data re-analysis and integration by applying it in a case study to data from deletion phenotype screens and mRNA expression profiling.

We have identified 141 novel palindromic repetitive elements in the genome of euryarchaeon Methanocaldococcus jannaschii. The total length of these elements is 14.3kb, which corresponds to 0.9% of the total genomic sequence and 6.3% of all extragenic regions. The elements can be divided into three groups (MJRE1-3) based on the sequence similarity. The low sequence identity within each of the groups suggests rather old origin of these elements in M. jannaschii. Three MJRE2 elements were located within the protein coding regions without disrupting the coding potential of the host genes, indicating that insertion of repeats might be a widespread mechanism to enhance sequence diversity in coding regions.

Seven distinct genome-wide divergence measures were applied pairwise to the nine sequenced animal genomes of human, mouse, rat, chicken, pufferfish, fruit fly, mosquito, and two nematode worms (Caenorhabditis briggsae and Caenorhabditis elegans). Qualitatively, all of these divergence measures are found to correlate with the estimated time since speciation; however, marked deviations are observed in a few lineages. The distinct genome divergence measures also correlate well among themselves, indicating that most of the processes shaping genomes are dominated by neutral events. The deviations from the clock-like scenario in some lineages are observed consistently by several measures, implicitly confirming their reliability.

Domains present one of the most useful levels at which to understand protein function, and domain family-based analysis has had a profound impact on the study of individual proteins. Protein domain discovery has been progressing steadily over the past 30 years. What are the realistically achievable goals of sequence-based domain analysis, and how far off are they for the sequences encoded in eukaryotic genomes? Here we address some of the issues involved in better coverage of sequence-based domain annotation, and the integration of these results within the wider context of genomes, structures and function.

The tuf gene of eubacteria, encoding the EF-tu elongation factor, was duplicated early in the evolution of the taxon. Phylogenetic and genomic location analysis of 20 complete eubacterial genomes suggests that this ancient duplication has been differentially lost and maintained in eubacteria.

We review data mining techniques in molecular biology, specifically those that extract information from the scientific literature itself. As more of the biological literature is published electronically, there is an opportunity, and even a need, to automatically summarize the literature in a customized way, for example by associating keywords to a topic. These keywords can be extracted from relevant publications. The process of keyword extraction can be automated and optimized to keep literature pointers automatically up-to-date or to filter relevant information from the literature. To illustrate these points, OMIM (Online Mendelian Inheritance in Man), a database of human inherited diseases, was linked to the literature and keywords were derived that covered distinct aspects such as genetic information on the one hand and disease-specific protein and phenotypic information on the other. They were used to extract information that is helpful for keeping entries about disease up-to-date.

Expressed sequence tag (EST) databases represent a large volume of information on expressed genes including tissue type, expression profile and exon structure. In this study we create an extensive data set of human alternative splicing. We report the analysis of 7867 non-redundant mRNAs, 3011 of which contained alternative splice forms (38% of all mRNAs analysed). From a total of 12572 ESTs 4560 different possible alternative splice forms were detected. Interestingly, 70% of the alternative splice forms correspond to exon deletion events with only 30% exonic insertions. We experimentally verified 19 different splice forms from 16 genes in a total subset of 20 studied; all of the respective genes are of medical relevance.

The flavoenzyme glutathione reductase catalyses electron transfer reactions between two major intracellular redox buffers, namely the NADPH/NADP+ couple and the 2 glutathione/glutathione disulfide couple. On this account, microcrystals of the enzyme were tested as redox probes of intracellular compartments. For introducing protein crystals into human fibroblasts, different methods (microinjection, particle bombardment and optical tweezers) were explored and compared. When glutathione reductase crystals are present in a cytosolic environment, the transition of the yellow Eox form to the orange-red 2-electron reduced charge transfer form, EH2, is observed. Taking into account the midpoint potential of the Eox/EH2 couple, the redox potential of the cytosol was found to be < -270 mV at pH 7.4 and 37 degrees C. As a general conclusion, competent proteins in crystalline--that is signal-amplifying--form are promising probes for studying intracellular events.

We introduce a simple and rapid strategy to identify genes that are responsible for species-specific phenotypes. The genome of a species that has a specific phenotype is compared with at least one, closely related, species that lacks this phenotype. Homologous genes that are shared among the species compared are identified and discarded from the list of candidates for species-specific genes. The process is automated and rapidly yields a small subset of the genome that likely contains genes responsible for the species-specific features. Functions are assigned to the genes, and dubious annotations are filtered out. Information is extracted not only from the presence of genes, but also from their absence with respect to known phenotypes. We have applied the technique to identify a set of species-specific genes in Helicobacter pylori by comparing it with its closest relatives for which complete genome sequences are available, Haemophilus influenzae and Escherichia coli. Of the genes of this set for which functional features can be obtained, a large fraction (63%, 123 proteins) is (potentially) involved in H. pylori's interaction with its host. We hypothesize that a family of outer membrane proteins is critical for the ability of H. pylori to colonize host cells in highly acidic environments.

New protein function is thought to evolve mostly by gene duplication and divergence. Here we present phylogenetic evidence that the multifunctional protein Fis of the gamma proteobacterial species derived from the COOH-terminal domain of an ancestral alpha proteobacterial NtrC transcriptional regulatory protein. All of the known enterobacterial fis genes are preceded by an open reading frame, named yhdG, that is highly similar to nifR3, a gene that forms an operon with ntrC in several alpha proteobacterial species. Thus, we propose that yhdG and fis were acquired by a lineage ancestral to the gamma proteobacteria in a single horizontal gene transfer event, and later diverged to their present functions.

We have identified, characterized and cloned human, mouse and chicken cDNA of a novel protein that binds to the Src homology domain 3 (SH3) of the Yes proto-oncogene product. We subsequently named it YAP for Yes-associated protein. Analysis of the YAP sequence revealed a protein module that was found in various structural, regulatory and signaling molecules. Because one of the prominent features of this sequence motif is the presence of two conserved tryptophans (W), we named it the WW domain. Using a functional screen of a cDNA expression library, we have identified two putative ligands of the WW domain of YAP which we named WBP-1 and WBP-2. Peptide sequence comparison between the two partial clones revealed a homologous proline-rich region. Binding assays and site-specific mutagenesis have shown that the proline-rich motif binds with relatively high affinity and specificity to the WW domain of YAP, with a preliminary consensus that is different from the SH3-binding PXXP motif. This suggests that the WW domain has a role in mediating protein-protein interactions via proline-rich regions, similar but distinct from Src homology 3 (SH3) domains. Based on this finding, we hypothesize that additional protein modules exist and that they could be isolated using proline-rich peptides as functional probes.

Recently, several groups have characterized and sequenced members of a new family of growth regulators (originally called cef10, connective tissue growth factor, fisp-12, cyr61, or, alternatively, beta IG-M1 and beta IG-M2), all of which belong to immediate-early genes expressed after induction by growth factors or certain oncogenes. Sequence analysis of this family revealed the presence of four distinct modules. Each module has homologues in other extracellular mosaic proteins such as Von Willebrand factor, slit, thrombospondins, fibrillar collagens, IGF-binding proteins and mucins. Classification and analysis of these modules suggests the location of binding regions and, by analogy to better characterized modules in other proteins, sheds some light onto the structure of this new family.

The sequence of RNase L has been re-examined by computer analysis. We propose a molecular architecture of RNase L, with an unusual combination, in one protein chain, of 9 ankyrin-like repeats, a functional active protein kinase and a C-terminal catalytic RNase similar to the yeast protein, IRE1. The protein kinase may be involved in a new signal transduction pathway which remains to be discovered.

Collagens are typical mosaic proteins containing a number of shuffled domains. These domains have been classified by sequence similarity in order to characterize their structural and functional relationships to other proteins. This analysis provides an overview of homologies of collagen domains. It also reveals two new relationships: (i) a module common to type V, IX, XI, and XII collagens was found to be homologous to the heparin binding domain of thrombospondin; (ii) the modular architecture of a human type VII collagen fragment was identified. Its N-terminal globular domain contains fibronectin type III repeats located adjacent to a Von Willebrand factor type A module. The proposed structural similarities point to analogous subfunctions of the respective domains in otherwise distinct proteins.

A new family of mosaic proteins is defined by sequence analysis. The family is characterized by a 260 residue domain common to proteins of apparently diverse function and tissue specificity: sperm receptors Zp2 and Zp3, betaglycan (also called TGF-beta type III receptor), uromodulin, as well as the major zymogen granule membrane protein (GP-2). The location of the common domain is similar with respect to putative transmembrane regions. The results lead to the hypothesis that this type of domain has a common tertiary structure and that there is a functional similarity in the recognition mechanism of the sperm receptor system and the TGF-beta receptor complex.

A comprehensive list of domains in extracellular mosaic proteins is presented. About 40 domains were distinguished by consensus patterns. A subsequent sequence database search recognized these domains in more than 200 extracellular proteins. The results point to a structural network, which may also represent the molecular basis for a complex coordination of various functions within the world of extracellular proteins.

Property patterns were constructed, based on an alignment of related domains in human complement subcomponents C1r and C1s as well as in the sea urchin protein uEGF. This kind of consensus pattern was able to identify similar domains in a human bone morphogenic protein, in a Xenopus laevis embryonal protein involved in dorsoanterior development and in a calcium-dependent serine protease secreted from malignant hamster embryo fibroblast cells. Because of the high level of overall sequence homology this protease may be the hamsters' equivalent of the human complement subcomponent C1s. The resulting multiple alignment of all studied domains suggests functionally and structurally important regions.

32 consensus patterns for a set of functional regions and structural motifs in protein sequences were constructed. The pattern definition is heuristic and based on 11 selected steric and physicochemical properties. By comparison with these patterns, it was possible to identify, without false detection, 1532 sites in 8702 protein sequences of SWISSPROT. Screening against such a pattern library offers a considerable chance to identify functional regions or structural motifs in proteins from which only the sequence is known.