BackgroundStructural variations (SVs) change the structure of the genome and are therefore the causes of various diseases. Next-generation sequencing allows us to obtain a multitude of sequence data, some of which can be used to infer the position of SVs.MethodsWe developed a new method and implementation named ClipCrop for detecting SVs with single-base resolution using soft-clipping information. A soft-clipped sequence is an unmatched fragment in a partially mapped read. To assess the performance of ClipCrop with other SV-detecting tools, we generated various patterns of simulation data – SV lengths, read lengths, and the depth of coverage of short reads – with insertions, deletions, tandem duplications, inversions and single nucleotide alterations in a human chromosome. For comparison, we selected BreakDancer, CNVnator and Pindel, each of which adopts a different approach to detect SVs, e.g. discordant pair approach, depth of coverage approach and split read approach, respectively.ResultsOur method outperformed BreakDancer and CNVnator in both discovering rate and call accuracy in any type of SV. Pindel offered a similar performance as our method, but our method crucially outperformed for detecting small duplications. From our experiments, ClipCrop infer reliable SVs for the data set with more than 50 bases read lengths and 20x depth of coverage, both of which are reasonable values in current NGS data set.ConclusionsClipCrop can detect SVs with higher discovering rate and call accuracy than any other tool in our simulation data set.

BackgroundMany learning approaches to predicting RNA-binding residues in a protein sequence construct a non-redundant training dataset based on the sequence similarity. The sequence similarity-based method either takes a whole sequence or discards it for a training dataset. However, similar sequences or even identical sequences can have different interaction sites depending on their interaction partners, and this information is lost when the sequences are removed. Furthermore, a training dataset constructed by the sequence similarity-based method may contain redundant data when the remaining sequence contains similar subsequences within the sequence. In addition to the problem with the training dataset, most approaches do not consider the interacting partner (i.e., RNA) of a protein when they predict RNA-binding amino acids. Thus, they always predict the same RNA-binding sites for a given protein sequence even if the protein binds to different RNA molecules.ResultsWe developed a feature vector-based method that removes data redundancy for a non-redundant training dataset. The feature vector-based method constructed a larger training dataset than the standard sequence similarity-based method, yet the dataset contained no redundant data. We identified effective features of protein and RNA (the interaction propensity of amino acid triplets, global features of the protein sequence, and RNA feature) for predicting RNA-binding residues. Using the method and features, we built a support vector machine (SVM) model that predicted RNA-binding residues in a protein sequence. Our SVM model showed an accuracy of 84.2%, an F-measure of 76.1%, and a correlation coefficient of 0.41 with 5-fold cross validation on a non-redundant dataset from 3,149 protein-RNA interacting pairs. In an independent test dataset that does not include the 3,149 pairs and were not used in training the SVM model, it achieved an accuracy of 90.3%, an F-measure of 72.8%, and a correlation coefficient of 0.24. Comparison with other methods on the same datasets demonstrated that our model was better than the others.ConclusionsThe feature vector-based redundancy reduction method is powerful for constructing a non-redundant training dataset for a learning model since it generates a larger dataset with non-redundant data than the standard sequence similarity-based method. Including the features of both RNA and protein sequences in a feature vector results in better performance than using the protein features only when predicting the RNA-binding residues in a protein sequence.

BackgroundThe Food and Drug Administration (FDA) approved drug labels contain a broad array of information, ranging from adverse drug reactions (ADRs) to drug efficacy, risk-benefit consideration, and more. However, the labeling language used to describe these information is free text often containing ambiguous semantic descriptions, which poses a great challenge in retrieving useful information from the labeling text in a consistent and accurate fashion for comparative analysis across drugs. Consequently, this task has largely relied on the manual reading of the full text by experts, which is time consuming and labor intensive.MethodIn this study, a novel text mining method with unsupervised learning in nature, called topic modeling, was applied to the drug labeling with a goal of discovering “topics” that group drugs with similar safety concerns and/or therapeutic uses together. A total of 794 FDA-approved drug labels were used in this study. First, the three labeling sections (i.e., Boxed Warning, Warnings and Precautions, Adverse Reactions) of each drug label were processed by the Medical Dictionary for Regulatory Activities (MedDRA) to convert the free text of each label to the standard ADR terms. Next, the topic modeling approach with latent Dirichlet allocation (LDA) was applied to generate 100 topics, each associated with a set of drugs grouped together based on the probability analysis. Lastly, the efficacy of the topic modeling was evaluated based on known information about the therapeutic uses and safety data of drugs.ResultsThe results demonstrate that drugs grouped by topics are associated with the same safety concerns and/or therapeutic uses with statistical significance (P<0.05). The identified topics have distinct context that can be directly linked to specific adverse events (e.g., liver injury or kidney injury) or therapeutic application (e.g., antiinfectives for systemic use). We were also able to identify potential adverse events that might arise from specific medications via topics.ConclusionsThe successful application of topic modeling on the FDA drug labeling demonstrates its potential utility as a hypothesis generation means to infer hidden relationships of concepts such as, in this study, drug safety and therapeutic use in the study of biomedical documents.

BackgroundBiologists require new algorithms to efficiently compress and store their large collections of phylogenetic trees. Our previous work showed that TreeZip is a promising approach for compressing phylogenetic trees. In this paper, we extend our TreeZip algorithm by handling trees with weighted branches. Furthermore, by using the compressed TreeZip file as input, we have designed an extensible decompressor that can extract subcollections of trees, compute majority and strict consensus trees, and merge tree collections using set operations such as union, intersection, and set difference.ResultsOn unweighted phylogenetic trees, TreeZip is able to compress Newick files in excess of 98%. On weighted phylogenetic trees, TreeZip is able to compress a Newick file by at least 73%. TreeZip can be combined with 7zip with little overhead, allowing space savings in excess of 99% (unweighted) and 92%(weighted). Unlike TreeZip, 7zip is not immune to branch rotations, and performs worse as the level of variability in the Newick string representation increases. Finally, since the TreeZip compressed text (TRZ) file contains all the semantic information in a collection of trees, we can easily filter and decompress a subset of trees of interest (such as the set of unique trees), or build the resulting consensus tree in a matter of seconds. We also show the ease of which set operations can be performed on TRZ files, at speeds quicker than those performed on Newick or 7zip compressed Newick files, and without loss of space savings.ConclusionsTreeZip is an efficient approach for compressing large collections of phylogenetic trees. The semantic and compact nature of the TRZ file allow it to be operated upon directly and quickly, without a need to decompress the original Newick file. We believe that TreeZip will be vital for compressing and archiving trees in the biological community.

BackgroundThe users desire to be provided short, specific answers to questions and put them in context by linking original sources from the biomedical literature. Through the use of information retrieval technologies, information systems retrieve information to index data based on all kinds of pre-defined searching techniques/functions such that various ranking strategies are designed depending on different sources. In this paper, we propose a robust approach to optimizing multi-source information for improving genomics retrieval performance.ResultsIn the proposed approach, we first consider a common scenario for a metasearch system that has access to multiple baselines with retrieving and ranking documents/passages by their own models. Then, given selected baselines from multiple sources, we investigate three modified fusion methods in the proposed approach, reciprocal, CombMNZ and CombSUM, to re-rank the candidates as the outputs for evaluation. Our empirical study on both 2007 and 2006 genomics data sets demonstrates the viability of the proposed approach for obtaining better performance. Furthermore, the experimental results show that the reciprocal method provides notable improvements on the individual baseline, especially on the passage2-level MAP and the aspect-level MAP.ConclusionsFrom the extensive experiments on two TREC genomics data sets, we draw the following conclusions. For the three fusion methods proposed in the robust approach, the reciprocal method outperforms the CombMNZ and CombSUM methods obviously, and CombSUM works well on the passage2-level when compared with CombMNZ. Based on the multiple sources of DFR, BM25 and language model, we can observe that the alliance of giants achieves the best result. Meanwhile, under the same combination, the better the baseline performance is, the more contribution the baseline provides. These conclusions are very useful to direct the fusion work in the field of biomedical information retrieval.

BackgroundDNA barcoding technology, which uses a short piece of DNA sequence to identify species, has wide ranges of applications. Until today, a universal DNA barcode marker for plants remains elusive. The rbc L and mat K regions have been proposed as the “core barcode” for plants and the ITS2 and psbA-trnH intergenic spacer (PTIGS) regions were later added as supplemental barcodes. The use of PTIGS region as a supplemental barcode has been limited by the lack of computational tools that can handle significant insertions and deletions in the PTIGS sequences. Here, we compared the most commonly used alignment-based and alignment-free methods and developed a web server to allow the biologists to carry out PTIGS-based DNA barcoding analyses.ResultsFirst, we compared several alignment-based methods such as BLAST and those calculating P distance and Edit distance, alignment-free methods Di-Nucleotide Frequency Profile (DNFP) and their combinations. We found that the DNFP and Edit-distance methods increased the identification success rate to ~80%, 20% higher than the most commonly used BLAST method. Second, the combined methods showed overall better success rate and performance. Last, we have developed a web server that allows (1) retrieving various sub-regions and the consensus sequences of PTIGS, (2) annotating novel PTIGS sequences, (3) determining species identity by PTIGS sequences using eight methods, and (4) examining identification efficiency and performance of the eight methods for various taxonomy groups.ConclusionsThe Edit distance and the DNFP methods have the highest discrimination powers. Hybrid methods can be used to achieve significant improvement in performance. These methods can be extended to applications using the core barcodes and the other supplemental DNA barcode ITS2. To our knowledge, the web server developed here is the only one that allows species determination based on PTIGS sequences. The web server can be accessed at http://psba-trnh-plantidit.dnsalias.org.