Summary: ORCAN (ORtholog sCANner) is a web-based meta-server for one-click evolutionary and functional annotation of protein sequences. The server combines information from the most popular orthology-prediction resources, including four tools and four online databases. Functional annotation utilizes five additional comparisons between the query and identified homologs, including: sequence similarity, protein domain architectures, functional motifs, Gene Ontology term assignments and a list of associated articles. Furthermore, the server uses a plurality-based rating system to evaluate the orthology relationships and to rank the reference proteins by their evolutionary and functional relevance to the query. Using a dataset of ∼1 million true yeast orthologs as a sample reference set, we show that combining multiple orthology-prediction tools in ORCAN increases the sensitivity and precision by 1-2 percent points. Availability and Implementation: The service is available for free at http://www.combio.pl/orcan/ . Contact: [email protected]. Supplementary information: Supplementary data are available at Bioinformatics online.

MOTIVATION: Argonaute-interacting WG/GW proteins are characterized by the presence of repeated sequence motifs containing glycine (G) and tryptophan (W). The motifs seem to be remarkably adaptive to amino acid substitutions and their sequences show non-contiguity. Our previous approach to the detection of GW domains, based on scoring their gross amino acid composition, allowed annotation of several novel proteins involved in gene silencing. The accumulation of new experimental data and more advanced applications revealed some deficiency of the algorithm in prediction selectivity. Additionally, W-motifs, though critical in gene regulation, have not yet been annotated in any available online resources. RESULTS: We present an improved set of computational tools allowing efficient management and annotation of W-based motifs involved in gene silencing. The new prediction algorithms provide novel functionalities by annotation of the W-containing domains at the local sequence motif level rather than by overall compositional properties. This approach represents a significant improvement over the previous method in terms of prediction sensitivity and selectivity. Application of the algorithm allowed annotation of a comprehensive list of putative Argonaute-interacting proteins across eukaryotes. An in-depth characterization of the domains' properties indicates its intrinsic disordered character. In addition, we created a knowledge-based portal (whub) that provides access to tools and information on RNAi-related tryptophan-containing motifs. AVAILABILITY AND IMPLEMENTATION: The web portal and tools are freely available at http://www.comgen.pl/whub. CONTACT: [email protected] SUPPLEMENTARY INFORMATION: Supplementary data are available at Bioinformatics online.

MOTIVATION: AGO(Argonaute)-binding domains, composed of repeated motifs, in which only binary combinations of tryptophan and glycine are conserved, bind AGO proteins and are essential during RNAi-mediated gene silencing. The amino acid sequence of this domain is extremely divergent and therefore very difficult to detect. Commonly used bioinformatic tools fail to identify tryptophan-glycine and/or glycine-tryptophan motifs (WG/GW) domains and currently there is no publicly available software which can detect these weakly conserved, but functional AGO-binding segments. RESULTS: Recently, we have developed an algorithm based on compositional analysis of the amino acid content of the domain. We have demonstrated that the algorithm can be successfully applied for the identification of the new WG/GW proteins in the Arabidopsis genome. Here we introduce Agos (Argonaute-binding domain screener), a novel universal web service for de novo identification of WG/GW domains in protein sequences. The web implementation of the algorithm contains several new features and enhancements: (i) one universal scoring matrix which allows identification of AGO-binding proteins in sequences representing all organisms; (ii) reduction of false positive predictions by improved selectivity of the algorithm; (iii) graphical interface to easily browse the prediction results; and (iv) the option to submit a DNA sequence which will be automatically translated in six frames before running the prediction algorithm. AVAILABILITY: Freely available at: http://bioinfo.amu.edu.pl/agos/.