共找到 20 条结果
In the early 1990s, after more than three decades of studying algorithms within the frame work of theoretical computer science, I shifted my focus to alogrithmic problems arising in genomics. There is a fundamental difference between the views of algorithms in the two fields: in theoretical computer science the input-output behavior of an algorithm is rigorously specified in advance, whereas in computational biology an algorithm is merely a vehicle for discovering Nature's ground truth. In order to be effective in computational genomics I have had to radically change my approach to research. On the occasion of this keynote address I will share some of the lessons I have learned, in the hope of making the way easier for computer scientists and mathematicians entering this field. These lessons will be encapsulated in a list of aphorisms, accompanied by illustrative examples.
Peptide-based vaccines, in which small peptides derived from target proteins (eptiopes) are used to provoke an immune reaction, have attracted considerable attention recently as a potential means both of treating infectious diseases and promoting the destruction of cancerous cells by a patient's own immune system. With the availability of large sequence databases and computers fast enough for rapid processing of large numbers of peptides, computer aided design of peptide-based vaccines has emerged as a promising approach to screening among billions of possible immune-active peptides to find those likely to provoke an immune response to a particular cell type. In this paper, we describe the development of three novel classes of methods for the prediction problem. We present a quadratic programming approach that can be trained on quantitative as well as qualitative data. The second method uses linear programming to counteract the fact that our training data contains mostly positive examples. The third class of methods uses sequence profiles obtained by clustering known epitopes to score candidate peptides. By integrating these methods, using a simple voting heuristic, we achieve improved accuracy over the state of the art.
The Cray MTA-2 (Multithreaded Architecture) is an unusual parallel supercomputer that promises ease of use and high performance. We describe our experience on the MTA-2 with a molecular dynamics code, SIMU-MD, that we are using to simulate the translocation of DNA through a nanopore in a silicon based ultrafast sequencer. Our sequencer is constructed using standard VLSI technology and consists of a nanopore surrounded by Field Effect Transistors (FETs). We propose to use the FETs to sense variations in charge as a DNA molecule translocates through the pore and thus differentiate between the four building block nucleotides of DNA. We were able to port SIMU-MD, a serial C code, to the MTA with only a modest effort and with good performance. Our porting process needed neither a parallelism support platform nor attention to the intimate details of parallel programming and interprocessor communication, as would have been the case with more conventional supercomputers.
We have constructed a computer model of the cytotoxic T lymphocyte (CTL) response to antigen and the maintenance of immunological memory. Because immune responses often begin with small numbers of cells and there is great variation among individual immune systems, we have chosen to implement a stochastic model that captures the life cycle of T cells more faithfully than deterministic models. Past models of the immune response have been differential equation based, which do not capture stochastic effects, or agent-based, which are computationally expensive. We use a stochastic stage-structured approach that has many of the advantages of agent-based modeling but is much more efficient. Our model can provide insights into the effect infections have on the CTL repertoire and the response to subsequent infections.
Analyzing protein sequence data becomes increasingly important recently. Most previous work on this area has mainly focused on building classification models. In this paper, we investigate in the problem of automatic clustering of unlabeled protein sequences. As a widely recognized technique in statistics and computer science, clustering has been proven very useful in detecting unknown object categories and revealing hidden correlations among objects. One difficulty that prevents clustering from being performed directly on protein sequence is the lack of an effective similarity measure that can be computed efficiently. Therefore, we propose a novel model for protein sequence cluster by exploring significant statistical properties possessed by the sequences. The concept of imprecise probabilities are introduced to the original probabilistic suffix tree to monitor the convergence of the empirical measurement and to guide the clustering process. It has been demonstrated that the proposed method can successfully discover meaningful families without the necessity of learning models of different families from pre-labeled "training data".
With the development of microarray techniques, there is an increasing need of information processing methods to analyze the high throughput data. Clustering is one of the most promising candidates because of its simplicity, flexibility and robustness. However, there is no "perfect" clustering approach outperforming its counterparts, and it is hard to evaluate and combine the results from different techniques, especially in a field without much prior knowledge, such as bioinformatics. This paper proposes a meta-clustering approach to extract the information from results of different clustering techniques, so that a better interpretation of the data distribution can be obtained. A special distance measure is defined to represent the statistical "signal" of each cluster produced by various clustering techniques. The algorithm is applied on both artificial and real data Simulations show that the proposed approach is able to extract the information efficiently and accurately from the input clustering structure.
NMR resonance assignment is one of the key steps in solving an NMR protein structure. The assignment process links resonance peaks to individual residues of the target protein sequence, providing the prerequisite for establishing intra- and inter-residue spatial relationships between atoms. The assignment process is tedious and time-consuming, which could take many weeks. Though there exist a number of computer programs to assist the assignment process, many NMR labs are still doing the assignments manually to ensure quality. This paper presents a new computational method based on our recent work towards automating the assignment process, particularly the process of backbone resonance peak assignment. We formulate the assignment problem as a constrained weighted bipartite matching problem. While the problem, in the most general situation, is NP-hard, we present an efficient solution based on a branch-and-bound algorithm with effective bounding techniques and a greedy filtering algorithm for reducing the search space. Our experimental results on 70 instances of (pseudo) real NMR data derived from 14 proteins demonstrate that the new solution runs much faster than a recently introduced (exhaustive) two-layer algorithm and recovers more correct peak assignments than the two-layer algorithm.
NMR resonance peak assignment is one of the key steps in solving an NMR protein structure. The assignment process links resonance peaks to individual residues of the target protein sequence, providing the prerequisite for establishing intra- and inter-residue spatial relationships between atoms. The assignment process is tedious and time-consuming, which could take many weeks. Though there exist a number of computer programs to assist the assignment process, many NMR labs are still doing the assignments manually to ensure quality. This paper presents (1) a new scoring system for mapping spin systems to residues, (2) an automated adjacency information extraction procedure from NMR spectra, and (3) a very fast assignment algorithm based on our previous proposed greedy filtering method and a maximum matching algorithm to automate the assignment process. The computational tests on 70 instances of (pseudo) experimental NMR data of 14 proteins demonstrate that the new score scheme has much better discerning power with the aid of adjacency information between spin systems simulated across various NMR spectra. Typically, with automated extraction of adjacency information, our method achieves nearly complete assignments for most of the proteins. The experiment shows very promising perspective that the fast automated assignment algorithm together with the new score scheme and automated adjacency extraction may be ready for practical use.
The personal information encoded in genomic DNA should not be made available to the public. With the increasing discoveries of new genes, it has become necessary to establish a security system for personal genome information. Although many security systems that are applied for electrical information in computers have been developed and established, there is no security system for information at DNA level. In this paper, we describe a new security system for information encoded within DNA. The original genomic DNA was mixed with many kinds of dummy DNAs (mixtures of natural and/or artificial DNAs) resulting in the masking of the original information. Using these dummy molecules, we succeeded to completely 'lock'the original genome information. If this information must be 'unlocked', it can be extracted and analyzed by a removal of dummy DNAs using molecular tagging techniques or by selective amplification using key primers.
Bioinformatics has become an active research area in recent years. The amount of mapped sequences doubles every fourteen months. BLAST has been widely employed for retrieving sequences which has similar portion(s) to a given sequence. However, BLAST has to scan the entire database every time when a query is issued. This can be very time consuming especially when the database is large. In this paper, we study the problem on how to build a persistent index structure for protein sequences to support approximate match. The suffix tree has been proposed as a solution to index sequence database and has been deployed on organizing DNA sequences (Hunt et al. 2001). Unfortunately, it suffers from the problem of "memory bottleneck" that prevents it from being applied efficiently to a large database. The performance even degrades further for protein database due to a larger fanout at each node. Here, we employ an indexing structure, called BASS-tree, to support approximate match in sublinear time on a large protein database. We call this indexing method as sequence approximate match (SAM) index method. The search of approximate matches can be properly directed to the portion in the database with a high potential of matching quickly. It has been demonstrated in our experiments that the potential performance improvement is in an order of magnitude over alternative methods such as the BLAST algorithm and the suffix tree.
Comparative analysis of syntenic genome sequences can be used to identify functional sites such as exons and regulatory elements. Here, the first step is to align two or several evolutionary related sequences and, in recent years, a number of computer programs have been developed for alignment of large genomic sequences. Some of these programs are extremely fast but often time-efficiency is achieved at the expense of sensitivity. One way of combining speed and sensitivity is to use an anchored-alignment approach. In a first step, a fast heuristic identifies a chain of strong sequence similarities that serve as anchor points. In a second step, regions between these anchor points are aligned using a slower but more sensitive method. We present CHAOS, a novel algorithm for rapid identification of chains of local sequence similarities among large genomic sequences. Similarities identified by CHAOS are used as anchor points to improve the running time of the DIALIGN alignment program. Systematic test runs show that this method can reduce the running time of DIALIGN by more than 93% while affecting the quality of the resulting alignments by only 1%. The source code for CHAOS is available at http://www.stanford.edu/~brudno/chaos/ An integrated program package containing CHAOS and DIALIGN is available at http://bibiserv.techfak.uni-bielefeld.de/dialign/
A new and essentially simple method to reconstruct prokaryotic phylogenetic trees from their complete genome data without using sequence alignment is proposed. It is based on the appearance frequency of oligopeptides of a fixed length (up to K = 6) in their proteomes. This is a method without fine adjustment and choice of genes. It can incorporate the effect of lateral gene transfer to some extent and leads to results comparable with the bacteriologists' systematics as reflected in the latest 2001 edition of the Bergey's Manual of Systematic Bacteriology [1, 2]. A key point in our approach is subtraction of a random background by using a Markovian model of order K - 1 from the composition vectors to highlight the shaping role of natural selection.
A simple statistical block code in combination with the LZW-based compression utilities gzip and compress has been found to increase by a significant amount the level of compression possible for the proteins encoded in Haemophilus influenzae, the first fully sequenced genome. The method yields an entropy value of 3.665 bits per symbol (bps), which is 0.657 bps below the maximum of 4.322 bps and an improvement of 0.452 bps over the best known to date of 4.118 bps using Matsumoto, Sadakane, and Imai's lza-CTW algorithm. Calculations based on a compact inverse genetic code show that the genome has a maximum entropy of 1.757 bps for the coding regions, with a possibly lower actual entropy. These results hint at the existence of hitherto unexplored redundancies that do not show up in Markov models and are indicative of more internal structure than suspected in both the protein and the genome.
In this paper we present a new Multiple Sequence Alignment (MSA) algorithm called AntiClusAl. The method makes use of the commonly use idea of aligning homologous sequences belonging to classes generated by some clustering algorithm, and then continue the alignment process ina bottom-up way along a suitable tree structure. The final result is then read at the root of the tree. Multiple sequence alignment in each cluster makes use of the progressive alignment with the 1-median (center) of the cluster. The 1-median of set S of sequences is the element of S which minimizes the average distance from any other sequence in S. Its exact computation requires quadratic time. The basic idea of our proposed algorithm is to make use of a simple and natural algorithmic technique based on randomized tournaments which has been successfully applied to large size search problems in general metric spaces. In particular a clustering algorithm called Antipole tree and an approximate linear 1-median computation are used. Our algorithm compared with Clustal W, a widely used tool to MSA, shows a better running time results with fully comparable alignment quality. A successful biological application showing high aminoacid conservation during evolution of Xenopus laevis SOD2 is also cited.
We propose a statistical method for estimating a gene network based on Bayesian networks from microarray gene expression data together with biological knowledge including protein-protein interactions, protein-DNA interactions, binding site information, existing literature and so on. Unfortunately, microarray data do not contain enough information for constructing gene networks accurately in many cases. Our method adds biological knowledge to the estimation method of gene networks under a Bayesian statistical framework, and also controls the trade-off between microarray information and biological knowledge automatically. We conduct Monte Carlo simulations to show the effectiveness of the proposed method. We analyze Saccharomyces cerevisiae gene expression data as an application.
A phylogenetic network is a generalization of a phylogenetic tree, allowing structural properties that are not tree-like. With the growth of genomic data, much of which does not fit ideal tree models, there is greater need to understand the algorithmics and combinatorics of phylogenetic networks [10, 11]. However, to date, very little has been published on this, with the notable exception of the paper by Wang et al.[12]. Other related papers include [4, 5, 7] We consider the problem introduced in [12], of determining whether the sequences can be derived on a phylogenetic network where the recombination cycles are node disjoint. In this paper, we call such a phylogenetic network a "galled-tree". By more deeply analysing the combinatorial constraints on cycle-disjoint phylogenetic networks, we obtain an efficient algorithm that is guaranteed to be both a necessary and sufficient test for the existence of a galled-tree for the data. If there is a galled-tree, the algorithm constructs one and obtains an implicit representation of all the galled trees for the data, and can create these in linear time for each one. We also note two additional results related to galled trees: first, any set of sequences that can be derived on a galled tree can be derived on a true tree (without recombination cycles), where at most one back mutation is allowed per site; second, the site compatibility problem (which is NP-hard in general) can be solved in linear time for any set of sequences that can be derived on a galled tree. The combinatorial constraints we develop apply (for the most part) to node-disjoint cycles in any phylogenetic network (not just galled-trees), and can be used for example to prove that a given site cannot be on a node-disjoint cycle in any phylogenetic network. Perhaps more important than the specific results about galled-trees, we introduce an approach that can be used to study recombination in phylogenetic networks that go beyond galled-trees.
The cWINNOWER algorithm detects fuzzy motifs in DNA sequences rich in protein-binding signals. A signal is defined as any short nucleotide pattern having up to d mutations differing from a motif of length l. The algorithm finds such motifs if multiple mutated copies of the motif (i.e., the signals) are present in the DNA sequence in sufficient abundance. The cWINNOWER algorithm substantially improves the sensitivity of the winnower method of Pevzner and Sze by imposing a consensus constraint, enabling it to detect much weaker signals. We studied the minimum number of detectable motifs qc as a function of sequence length N for random sequences. We found that q(c) increases linearly with N for a fast version of the algorithm based on counting three-member sub-cliques. Imposing consensus constraints reduces q(c) by a factor of three in this case, which makes the algorithm dramatically more sensitive. Our most sensitive algorithm, which counts four-member sub-cliques, needs a minimum of only 13 signals to detect motifs in a sequence of length N = 12,000 for (l,d) = (15,4).
Nowadays, high throughput experimental techniques make it feasible to examine and collect massive data at the molecular level. These data, typically mapped to a very high dimensional feature space, carry rich information about functionalities of certain chemical or biological entities and can be used to infer valuable knowledge for the purposes of classification and prediction. Typically, a small number of features or feature combinations may play determinant roles in functional discrimination. The identification of such features or feature combinations is of great importance. In this paper, we study the problem of discovering compact and highly discriminative features or feature combinations from a rich feature collection. We employ the support vector machine as the classification means and aim at finding compact feature combinations. Comparing to previous methods on feature selection, which identify features solely based on their individual roles in the classification, our method is able to identify minimal feature combinations that ultimately have determinant roles in a systematic fashion. Experimental study on drug activity data shows that our method can discover descriptors that are not necessarily significant individually but are most significant collectively.