Proceedings. IEEE Computational Systems Bioinformatics Conference最新文献

The inverse protein folding problem is that of designing an amino acid sequence which has a particular native protein fold. This problem arises in drug design where a particular structure is necessary to ensure proper protein-protein interactions. In this paper we show that in the 2D HP model of Dill it is possible to solve this problem for a broad class of structures. These structures can be used to closely approximate any given structure. One of the most important properties of a good protein is its stability -- the aptitude not to fold simultanously into other structures. We show that for a number of basic structures, our sequences have a unique fold.

RNA structures can be viewed as a kind of special strings with some characters bonded with each other. The question of aligning two RNA structures has been studied for a while, and there are several successful algorithms that are based upon different models. In this paper, by adopting the model introduced in [18], we propose two algorithms to attack the question of aligning multiple RNA structures. We reduce the multiple RNA structure alignment problem to the problem of aligning two RNA structure alignments.

With the recent explosion of interest in microarray technology, massive amounts of microarray images are currently being produced. The storage and the transmission of this type of data are becoming increasingly challenging. Here we propose lossless and lossy compression algorithms for microarray images originally digitized at 16 bpp (bits per pixels) that achieve an average of 9.5 - 11.5 bpp (lossless) and 4.6 - 6.7 bpp (lossy, with a PSNR of 63 dB). The lossy compression is applied only on the background of the image, thereby preserving the regions of interest. The methods are based on a completely automatic gridding procedure of the image.

We propose derivative Boyer-Moore (d-BM), a new compressed pattern matching algorithm in DNA sequences. This algorithm is based on the Boyer-Moore method, which is one of the most popular string matching algorithms. In this approach, we compress both DNA sequences and patterns by using two bits to represent each A, T, C, G character. Experiments indicate that this compressed pattern matching algorithm searches long DNA patterns (length > 50) more than 10 times faster than the exact match routine of the software package Agrep, which is known as the fastest pattern matching tool. Moreover, compression of DNA sequences by this method gives a guaranteed space saving of 75%. In part the enhanced speed of the algorithm is due to the increased efficiency of the Boyer-Moore method resulting from an increase in alphabet size from 4 to 256.

Motivation: Since the whole genome sequences for many species are currently available, computational predictions of RNA secondary structures and computational identifications of those non-coding RNA regions by comparative genomics become important, and require more advanced alignment methods. Recently, an approach of structural alignments for RNA sequences has been introduced to solve these problems. By structural alignments, we mean a pairwise alignment to align an unfolded RNA sequence into a folded RNA sequence of known secondary structure. Pair HMMs on tree structures (PHMMTSs) proposed by Sakakibara are efficient automata-theoretic models for structural alignments of RNA secondary structures, but are incapable of handling pseudoknots. On the other hand, tree adjoining grammars (TAGs) is a subclass of context-sensitive grammar, which is suitable for modeling pseudoknots. Our goal is to extend PHMMTSs by incorporating TAGs to be able to handle pseudoknots.

Results: We propose the pair stochastic tree adjoining grammars (PSTAGs) for modeling RNA secondary structures including pseudoknots and show the strong experimental evidences that modeling pseudoknot structures significantly improves the prediction accuracies of RNA secondary structures. First, we extend the notion of PHMMTSs defined on alignments of 'trees' to PSTAGs defined on alignments of "TAG (derivation) trees", which represent a top-down parsing process of TAGs and are functionally equivalent to derived trees of TAGs. Second, we modify PSTAGs so that it takes as input a pair of a linear sequence and a TAG tree representing a pseudoknot structure of RNA to produce a structural alignment. Then, we develop a polynomial-time algorithm for obtaining an optimal structural alignment by PSTAGs, based on dynamic programming parser. We have done several computational experiments for predicting pseudoknots by PSTAGs, and our computational experiments suggests that prediction of RNA pseudoknot structures by our method are more efficient and biologically plausible than by other conventional methods. The binary code for PSTAG method is freely available from our website at http://www.dna.bio.keio.ac.jp/pstag/.

Phylogenetic trees are commonly reconstructed based on hard optimization problems such as maximum parsimony (MP) and maximum likelihood (ML). Conventional MP heuristics for producing phylogenetic trees produce good solutions within reasonable time on small datasets (up to a few thousand sequences), while ML heuristics are limited to smaller datasets (up to a few hundred sequences). However, since MP (and presumably ML) is NP-hard, such approaches do not scale when applied to large datasets. In this paper, we present a new technique called Recursive-Iterative-DCM3 (Rec-I-DCM3), which belongs to our family of Disk-Covering Methods (DCMs). We tested this new technique on ten large biological datasets ranging from 1,322 to 13,921 sequences and obtained dramatic speedups as well as significant improvements in accuracy (better than 99.99%) in comparison to existing approaches. Thus, high-quality reconstructions can be obtained for datasets at least ten times larger than was previously possible.

A great deal of recent research has focused on the challenging task of selecting differentially expressed genes from microarray data ('gene selection'). Numerous gene selection algorithms have been proposed in the literature, but it is often unclear exactly how these algorithms respond to conditions like small sample-sizes or differing variances. Choosing an appropriate algorithm can therefore be difficult in many cases. In this paper we propose a theoretical analysis of gene selection, in which the probability of successfully selecting relevant genes, using a given gene ranking function, is explicitly calculated in terms of population parameters. The theory developed is applicable to any ranking function which has a known sampling distribution, or one which can be approximated analytically. In contrast to empirical methods, the analysis can easily be used to examine the behaviour of gene selection algorithms under a wide variety of conditions, even when the numbers of genes involved runs into the tens of thousands. The utility of our approach is illustrated by comparing three well-known gene ranking functions.

The discovery of novel non-coding RNAs has been among the most exciting recent developments in Biology. Yet, many more remain undiscovered. It has been hypothesized that there is in fact an abundance of functional non-coding RNA (ncRNA) with various catalytic and regulatory functions. Computational methods tailored specifically for ncRNA are being actively developed. As the inherent signal for ncRNA is weaker than that for protein coding genes, comparative methods offer the most promising approach, and are the subject of our research. We consider the following problem: Given an RNA sequence with a known secondary structure, efficiently compute all structural homologs (computed as a function of sequence and structural similarity) in a genomic database. Our approach, based on structural filters that eliminate a large portion of the database, while retaining the true homologs allows us to search a typical bacterial database in minutes on a standard PC, with high sensitivity and specificity. This is two orders of magnitude better than current available software for the problem.

The soundness of clustering in the analysis of gene expression profiles and gene function prediction is based on the hypothesis that genes with similar expression profiles may imply strong correlations with their functions in the biological activities. Gene Ontology (GO) has become a well accepted standard in organizing gene function categories. Different gene function categories in GO can have very sophisticated relationships, such as 'part of' and 'overlapping'. Until now, no clustering algorithm can generate gene clusters within which the relationships can naturally reflect those of gene function categories in the GO hierarchy. The failure in resembling the relationships may reduce the confidence of clustering in gene function prediction. In this paper, we present a new clustering technique, Smart Hierarchical Tendency Preserving clustering (SHTP-clustering), based on a bicluster model, Tendency Preserving cluster (TP-Cluster). By directly incorporating Gene Ontology information into the clustering process, the SHTP-clustering algorithm yields a TP-cluster tree within which any subtree can be well mapped to a part of the GO hierarchy. Our experiments on yeast cell cycle data demonstrate that this method is efficient and effective in generating the biological relevant TP-Clusters.

Researchers, hindered by a lack of standard gene and protein-naming conventions, endure long, sometimes fruitless, literature searches. A system is described which is able to automatically assign gene names to their LocusLink ID (LLID) in previously unseen MEDLINE abstracts. The system is based on supervised learning and builds a model for each LLID. The training sets for all LLIDs are extracted automatically from MEDLINE references in the LocusLink and SwissProt databases. A validation was done of the performance for all 20,546 human genes with LLIDs. Of these, 7,344 produced good quality models (F-measure > 0.7, nearly 60% of which were > 0.9) and 13,202 did not, mainly due to insufficient numbers of known document references. A hand validation of MEDLINE documents for a set of 66 genes agreed well with the system's internal accuracy assessment. It is concluded that it is possible to achieve high quality gene disambiguation using scaleable automated techniques.