BACKGROUND: DNA methylation is an important epigenetic modification critical in regulation and transgenerational inheritance. The methylation level can be estimated at single-nucleotide resolution by whole-genome bisulfite sequencing (BS-seq; WGBS). Current bisulfite aligners provide pipelines for processing the reads by WGBS; however, few are able to analyze the BS-seqs in a reasonable timeframe that meets the needs of the rapid expansion of epigenome sequencing in biomedical research. RESULTS: We introduce BS-Seeker3, an extensively improved and optimized implementation of BS-Seeker2 that leverages the available computational power of a standard bioinformatics lab. BS-Seeker3 adopts all alignment features of BS-Seeker2. It performs ultrafast alignments and achieves both high accuracy and high mappability, more than twice that of the other aligners that we evaluated. Moreover, BS Seeker 3 is well linked with downstream analyzer MethGo for up to 9 types of genomic and epigenomic analyses. CONCLUSIONS: BS-Seeker3 is an accurate, versatile, ultra-fast pipeline for processing bisulfite-converted reads. It also helps the user better visualize the methylation data.

BACKGROUND: Bisulfite sequencing using next generation sequencers yields genome-wide measurements of DNA methylation at single nucleotide resolution. Traditional aligners are not designed for mapping bisulfite-treated reads, where the unmethylated Cs are converted to Ts. We have developed BS Seeker, an approach that converts the genome to a three-letter alphabet and uses Bowtie to align bisulfite-treated reads to a reference genome. It uses sequence tags to reduce mapping ambiguity. Post-processing of the alignments removes non-unique and low-quality mappings. RESULTS: We tested our aligner on synthetic data, a bisulfite-converted Arabidopsis library, and human libraries generated from two different experimental protocols. We evaluated the performance of our approach and compared it to other bisulfite aligners. The results demonstrate that among the aligners tested, BS Seeker is more versatile and faster. When mapping to the human genome, BS Seeker generates alignments significantly faster than RMAP and BSMAP. Furthermore, BS Seeker is the only alignment tool that can explicitly account for tags which are generated by certain library construction protocols. CONCLUSIONS: BS Seeker provides fast and accurate mapping of bisulfite-converted reads. It can work with BS reads generated from the two different experimental protocols, and is able to efficiently map reads to large mammalian genomes. The Python program is freely available at http://pellegrini.mcdb.ucla.edu/BS_Seeker/BS_Seeker.html.

Protein interactions play a vital part in the function of a cell. As experimental techniques for detection and validation of protein interactions are time consuming, there is a need for computational methods for this task. Protein interactions appear to form a network with a relatively high degree of local clustering. In this paper we exploit this clustering by suggesting a score based on triplets of observed protein interactions. The score utilises both protein characteristics and network properties. Our score based on triplets is shown to complement existing techniques for predicting protein interactions, outperforming them on data sets which display a high degree of clustering. The predicted interactions score highly against test measures for accuracy. Compared to a similar score derived from pairwise interactions only, the triplet score displays higher sensitivity and specificity. By looking at specific examples, we show how an experimental set of interactions can be enriched and validated. As part of this work we also examine the effect of different prior databases upon the accuracy of prediction and find that the interactions from the same kingdom give better results than from across kingdoms, suggesting that there may be fundamental differences between the networks. These results all emphasize that network structure is important and helps in the accurate prediction of protein interactions. The protein interaction data set and the program used in our analysis, and a list of predictions and validations, are available at http://www.stats.ox.ac.uk/bioinfo/resources/PredictingInteractions.

MOTIVATION: The Majority Vote approach has demonstrated that protein-protein interactions can be used to predict the structure or function of a protein. In this article we propose a novel method for the prediction of such protein characteristics based on frequencies of pairwise interactions. In addition, we study a second new approach using the pattern frequencies of triplets of proteins, thus for the first time taking network structure explicitly into account. Both these methods are extended to jointly consider multiple organisms and multiple characteristics. RESULTS: Compared to the standard non-network-based method, namely the Majority Vote method, in large networks our predictions tend to be more accurate. For structure prediction, the Frequency-based method reaches up to 71% accuracy, and the Triplet-based method reaches up to 72% accuracy, whereas for function prediction, both the Triplet-based method and the Frequency-based method reach up to 90% accuracy. Function prediction on proteins without homologues showed slightly less but comparable accuracies. Including partially annotated proteins substantially increases the number of proteins for which our methods predict their characteristics with reasonable accuracy. We find that the enhanced Triplet-based method does not currently yield significantly better results than the enhanced Frequency-based method, suggesting that triplets of interactions do not contain substantially more information about protein characteristics than interaction pairs. Our methods offer two main improvements over current approaches--first, multiple protein characteristics are considered simultaneously, and second, data is integrated from multiple species. In addition, the Triplet-based method includes network structure more explicitly than the Majority Vote and the Frequency-based method. AVAILABILITY: The program is available upon request. SUPPLEMENTARY INFORMATION: Supplementary data are available at Bioinformatics online.

BACKGROUND: Proteins control and mediate many biological activities of cells by interacting with other protein partners. This work presents a statistical model to predict protein interaction networks of Drosophila melanogaster based on insight into domain interactions. RESULTS: Three high-throughput yeast two-hybrid experiments and the collection in FlyBase were used as our starting datasets. The co-occurrences of domains in these interactive events are converted into a probability score of domain-domain interaction. These scores are used to infer putative interaction among all available open reading frames (ORFs) of fruit fly. Additionally, the likelihood function is used to estimate all potential protein-protein interactions. All parameters are successfully iterated and MLE is obtained for each pair of domains. Additionally, the maximized likelihood reaches its converged criteria and maintains the probability stable. The hybrid model achieves a high specificity with a loss of sensitivity, suggesting that the model may possess major features of protein-protein interactions. Several putative interactions predicted by the proposed hybrid model are supported by literatures, while experimental data with a low probability score indicate an uncertain reliability and require further proof of interaction.Fly-DPI is the online database used to present this work. It is an integrated proteomics tool with comprehensive protein annotation information from major databases as well as an effective means of predicting protein-protein interactions. As a novel search strategy, the ping-pong search is a naïve path map between two chosen proteins based on pre-computed shortest paths. Adopting effective filtering strategies will facilitate researchers in depicting the bird's eye view of the network of interest. Fly-DPI can be accessed at http://flydpi.nhri.org.tw. CONCLUSION: This work provides two reference systems, statistical and biological, to evaluate the reliability of protein interaction. First, the hybrid model statistically estimates both experimental and predicted protein interaction relationships. Second, the biological information for filtering and annotation itself is a strong indicator for the reliability of protein-protein interaction. The space-temporal or stage-specific expression patterns of genes are also critical for identifying proteins involved in a particular situation.

We implemented a statistical model into our protein interaction database for validation of two-hybrid assays of Helicobacter pylori, and prediction of putative protein interactions not yet discovered experimentally. To present the enormous amount of experimental and inferred protein interaction networking maps, the H.pylori Database of Protein Interactomes (hp-DPI) is developed with a succinct yet comprehensive visualization tool integrated with annotation from Genbank, GO, and KEGG. hp-DPI is first built with, but not limited to, H.pylori protein interactions and is expected to naturally include other organisms' protein interacting relationships in the future.