Subtle motifs: defining the limits of motif finding algorithms

MOTIVATION: What constitutes a subtle motif? Intuitively, it is a motif that is almost indistinguishable, in the statistical sense, from random motifs. This question has important practical consequences: consider, for example, a biologist that is generating a sample of upstream regulatory sequences with the goal of finding a regulatory pattern that is shared by these sequences. If the sequences are too short then one risks losing some of the regulatory patterns that are located further upstream. Conversely, if the sequences are too long, the motif becomes too subtle and one is then likely to encounter random motifs which are at least as significant statistically as the regulatory pattern itself. In practical terms one would like to recognize the sequence length threshold, or the twilight zone, beyond which the motifs are in some sense too subtle. RESULTS: The paper defines the motif twilight zone where every motif finding algorithm would be exposed to random motifs which are as significant as the one which is sought. We also propose an objective tool for evaluating the performance of subtle motif finding algorithms. Finally we apply these tools to evaluate the success of our MULTIPROFILER algorithm to detect subtle motifs.     

Finding composite regulatory patterns in DNA sequences

Pattern discovery in unaligned DNA sequences is a fundamental problem in computational biology with important applications in finding regulatory signals. Current approaches to pattern discovery focus on monad patterns that correspond to relatively short contiguous strings. However, many of the actual regulatory signals are composite patterns that are groups of monad patterns that occur near each other. A difficulty in discovering composite patterns is that one or both of the component monad patterns in the group may be 'too weak'. Since the traditional monad-based motif finding algorithms usually output one (or a few) high scoring patterns, they often fail to find composite regulatory signals consisting of weak monad parts. In this paper, we present a MITRA (MIsmatch TRee Algorithm) approach for discovering composite signals. We demonstrate that MITRA performs well for both monad and composite patterns by presenting experiments over biological and synthetic data.     

Apples to apples: improving the performance of motif finders and their significance analysis in the Twilight Zone

MOTIVATION: Effective algorithms for finding relatively weak motifs are an important practical necessity while scanning long DNA sequences for regulatory elements. The success of such an algorithm hinges on the ability of its scoring function combined with a significance analysis test to discern real motifs from random noise. RESULTS: In the first half of the paper we show that the paradigm of relying on entropy scores and their E-values can lead to undesirable results when searching for weak motifs and we offer alternate approaches to analyzing the significance of motifs. In the second half of the paper we reintroduce a scoring function and present a motif-finder that optimizes it that are more effective in finding relatively weak motifs than other tools. AVAILABILITY: The GibbsILR motif finder is available at http://www.cs.cornell.edu/~keich.     

RankMotif++: a motif-search algorithm that accounts for relative ranks of K-mers in binding transcription factors

MOTIVATION: The sequence specificity of DNA-binding proteins is typically represented as a position weight matrix in which each base position contributes independently to relative affinity. Assessment of the accuracy and broad applicability of this representation has been limited by the lack of extensive DNA-binding data. However, new microarray techniques, in which preferences for all possible K-mers are measured, enable a broad comparison of both motif representation and methods for motif discovery. Here, we consider the problem of accounting for all of the binding data in such experiments, rather than the highest affinity binding data. We introduce the RankMotif++, an algorithm designed for finding motifs whenever sequences are associated with a semi-quantitative measure of protein-DNA-binding affinity. RankMotif++ learns motif models by maximizing the likelihood of a set of binding preferences under a probabilistic model of how sequence binding affinity translates into binding preference observations. Because RankMotif++ makes few assumptions about the relationship between binding affinity and the semi-quantitative readout, it is applicable to a wide variety of experimental assays of DNA-binding preference. RESULTS: By several criteria, RankMotif++ predicts binding affinity better than two widely used motif finding algorithms (MDScan, MatrixREDUCE) or more recently developed algorithms (PREGO, Seed and Wobble), and its performance is comparable to a motif model that separately assigns affinities to 8-mers. Our results validate the PWM model and provide an approximation of the precision and recall that can be expected in a genomic scan. AVAILABILITY: RankMotif++ is available upon request. SUPPLEMENTARY INFORMATION: Supplementary data are available at Bioinformatics online.     

A random graph approach to NMR sequential assignment.

Nuclear magnetic resonance (NMR) spectroscopy allows scientists to study protein structure, dynamics and interactions in solution. A necessary first step for such applications is determining the resonance assignment, mapping spectral data to atoms and residues in the primary sequence. Automated resonance assignment algorithms rely on information regarding connectivity (e.g., through-bond atomic interactions) and amino acid type, typically using the former to determine strings of connected residues and the latter to map those strings to positions in the primary sequence. Significant ambiguity exists in both connectivity and amino acid type information. This paper focuses on the information content available in connectivity alone and develops a novel random-graph theoretic framework and algorithm for connectivity-driven NMR sequential assignment. Our random graph model captures the structure of chemical shift degeneracy, a key source of connectivity ambiguity. We then give a simple and natural randomized algorithm for finding optimal assignments as sets of connected fragments in NMR graphs. The algorithm naturally and efficiently reuses substrings while exploring connectivity choices; it overcomes local ambiguity by enforcing global consistency of all choices. By analyzing our algorithm under our random graph model, we show that it can provably tolerate relatively large ambiguity while still giving expected optimal performance in polynomial time. We present results from practical applications of the algorithm to experimental datasets from a variety of proteins and experimental set-ups. We demonstrate that our approach is able to overcome significant noise and local ambiguity in identifying significant fragments of sequential assignments.     

Consensus higher order repeats and frequency of string distributions in human genome

Key string algorithm (KSA) could be viewed as robust computational generalization of restriction enzyme method. KSA enables robust and effective identification and structural analyzes of any given genomic sequences, like in the case of NCBI assembly for human genome. We have developed a method, using total frequency distribution of all r-bp key strings in dependence on the fragment length l, to determine the exact size of all repeats within the given genomic sequence, both of monomeric and HOR type. Subsequently, for particular fragment lengths equal to each of these repeat sizes we compute the partial frequency distribution of r-bp key strings; the key string with highest frequency is a dominant key string, optimal for segmentation of a given genomic sequence into repeat units. We illustrate how a wide class of 3-bp key strings leads to a key-string-dependent periodic cell which enables a simple identification and consensus length determinations of HORs, or any other highly convergent repeat of monomeric or HOR type, both tandem or dispersed. We illustrated KSA application for HORs in human genome and determined consensus HORs in the Build 35.1 assembly. In the next step we compute suprachromosomal family classification and CENP-B box / pJalpha distributions for HORs. In the case of less convergent repeats, like for example monomeric alpha satellite (20-40% divergence), we searched for optimal compact key string using frequency method and developed a concept of composite key string (GAAAC--CTTTG) or flexible relaxation (28 bp key string) which provides both monomeric alpha satellites as well as alpha monomer segmentation of internal HOR structure. This method is convenient also for study of R-strand (direct) / S-strand (reverse complement) alpha monomer alternations. Using KSA we identified 16 alternating regions of R-strand and S-strand monomers in one contig in choromosome 7. Use of CENP-B box and/or pJalpha motif as key string is suitable both for identification of HORs and monomeric pattern as well as for studies of CENP-B box / pJalpha distribution. As an example of application of KSA to sequences outside of HOR regions we present our finding of a tandem with highly convergent 3434-bp Long monomer in chromosome 5 (divergence less then 0.3%).     

Regulatory sequence analysis tools

The web resource Regulatory Sequence Analysis Tools (RSAT) (http://rsat.ulb.ac.be/rsat) offers a collection of software tools dedicated to the prediction of regulatory sites in non-coding DNA sequences. These tools include sequence retrieval, pattern discovery, pattern matching, genome-scale pattern matching, feature-map drawing, random sequence generation and other utilities. Alternative formats are supported for the representation of regulatory motifs (strings or position-specific scoring matrices) and several algorithms are proposed for pattern discovery. RSAT currently holds >100 fully sequenced genomes and these data are regularly updated from GenBank.     

Rapid motif-based prediction of circular permutations in multi-domain proteins

MOTIVATION: Rearrangements of protein domains and motifs such as swaps and circular permutations (CPs) can produce erroneous results in searching sequence databases when using traditional methods based on linear sequence alignments. Circular permutations are also of biological relevance because they can help to better understand both protein evolution and functionality. RESULTS: We have developed an algorithm, RASPODOM, which is based on the classical recursive alignment scheme. Sequences are represented as strings of domains taken from precompiled resources of domain (motif) databases such as ProDom. The algorithm works several orders of magnitude faster than a reimplementation of the existing CP detection algorithm working on strings of amino acids, produces virtually no false positives and allows the discrimination of true CPs from 'intermediate' CPs (iCPs). Several true CPs which have not been reported in literature so far could be identified from Swiss-Prot/TrEMBL within minutes.     

Mining frequent stem patterns from unaligned RNA sequences

MOTIVATION: In detection of non-coding RNAs, it is often necessary to identify the secondary structure motifs from a set of putative RNA sequences. Most of the existing algorithms aim to provide the best motif or few good motifs, but biologists often need to inspect all the possible motifs thoroughly. RESULTS: Our method RNAmine employs a graph theoretic representation of RNA sequences and detects all the possible motifs exhaustively using a graph mining algorithm. The motif detection problem boils down to finding frequently appearing patterns in a set of directed and labeled graphs. In the tasks of common secondary structure prediction and local motif detection from long sequences, our method performed favorably both in accuracy and in efficiency with the state-of-the-art methods such as CMFinder. AVAILABILITY: The software is available upon request.     

DNA sequence and structure: direct and indirect recognition in protein-DNA binding

MOTIVATION: Direct recognition, or direct readout, of DNA bases by a DNA-binding protein involves amino acids that interact directly with features specific to each base. Experimental evidence also shows that in many cases the protein achieves partial sequence specificity by indirect recognition, i.e., by recognizing structural properties of the DNA. (1) Could threading a DNA sequence onto a crystal structure of bound DNA help explain the indirect recognition component of sequence specificity? (2) Might the resulting pure-structure computational motif manifest itself in familiar sequence-based computational motifs? RESULTS: The starting structure motif was a crystal structure of DNA bound to the integration host factor protein (IHF) of E. coli. IHF is known to exhibit both direct and indirect recognition of its binding sites. (1) Threading DNA sequences onto the crystal structure showed statistically significant partial separation of 60 IHF binding sites from random and intragenic sequences and was positively correlated with binding affinity. (2) The crystal structure was shown to be equivalent to a linear Markov network, and so, to a joint probability distribution over sequences, computable in linear time. It was transformed algorithmically into several common pure-sequence representations, including (a) small sets of short exact strings, (b) weight matrices, (c) consensus regular patterns, (d) multiple sequence alignments, and (e) phylogenetic trees. In all cases the pure-sequence motifs retained statistically significant partial separation of the IHF binding sites from random and intragenic sequences. Most exhibited positive correlation with binding affinity. The multiple alignment showed some conserved columns, and the phylogenetic tree partially mixed low-energy sequences with IHF binding sites but separated high-energy sequences. The conclusion is that deformation energy explains part of indirect recognition, which explains part of IHF sequence-specific binding.     

Efficient estimation of graphlet frequency distributions in protein-protein interaction networks

MOTIVATION: Algorithmic and modeling advances in the area of protein-protein interaction (PPI) network analysis could contribute to the understanding of biological processes. Local structure of networks can be measured by the frequency distribution of graphlets, small connected non-isomorphic induced subgraphs. This measure of local structure has been used to show that high-confidence PPI networks have local structure of geometric random graphs. Finding graphlets exhaustively in a large network is computationally intensive. More complete PPI networks, as well as PPI networks of higher organisms, will thus require efficient heuristic approaches. RESULTS: We propose two efficient and scalable heuristics for finding graphlets in high-confidence PPI networks. We show that both PPI and their model geometric random networks, have defined boundaries that are sparser than the 'inner parts' of the networks. In addition, these networks exhibit 'uniformity' of local structure inside the networks. Our first heuristic exploits these two structural properties of PPI and geometric random networks to find good estimates of graphlet frequency distributions in these networks up to 690 times faster than the exhaustive searches. Our second heuristic is a variant of a more standard sampling technique and it produces accurate approximate results up to 377 times faster than the exhaustive searches. We indicate how the combination of these approaches may result in an even better heuristic. AVAILABILITY: Supplementary information is available at http://www.cs.toronto.edu/~natasha/BIOINF-2005-0946/Supplementary.pdf. Software implementing the algorithms is available at http://www.cs.toronto.edu/~natasha/BIOINF-2005-0946/estimate_grap-hlets.html. SUPPLEMENTARY INFORMATION: Supplementary data are available at Bioinformatics online.     

Fast exact algorithms for the closest string and substring problems with application to the planted (L, d)-motif model

We present two parameterized algorithms for the closest string problem. The first runs in O(nL + nd · 17.97d) time for DNA strings and in O(nL + nd · 61.86d) time for protein strings, where n is the number of input strings, L is the length of each input string, and d is the given upper bound on the number of mismatches between the center string and each input string. The second runs in O(nL + nd · 13.92d) time for DNA strings and in O(nL + nd · 47.21d) time for protein strings. We then extend the first algorithm to a new parameterized algorithm for the closest substring problem that runs in O((n - 1)m2(L + d · 17.97d · m[log2(d+1)])) time for DNA strings and in O((n - 1)m2(L + d · 61.86d · m[log2(d+1)])) time for protein strings, where n is the number of input strings, L is the length of the center substring, L - 1 + m is the maximum length of a single input string, and d is the given upper bound on the number of mismatches between the center substring and at least one substring of each input string. All the algorithms significantly improve the previous bests. To verify experimentally the theoretical improvements in the time complexity, we implement our algorithm in C and apply the resulting program to the planted (L, d)-motif problem proposed by Pevzner and Sze in 2000. We compare our program with the previously best exact program for the problem, namely PMSPrune (designed by Davila et al. in 2007). Our experimental data show that our program runs faster for practical cases and also for several challenging cases. Our algorithm uses less memory too.     

Inference from clustering with application to gene-expression microarrays.

There are many algorithms to cluster sample data points based on nearness or a similarity measure. Often the implication is that points in different clusters come from different underlying classes, whereas those in the same cluster come from the same class. Stochastically, the underlying classes represent different random processes. The inference is that clusters represent a partition of the sample points according to which process they belong. This paper discusses a model-based clustering toolbox that evaluates cluster accuracy. Each random process is modeled as its mean plus independent noise, sample points are generated, the points are clustered, and the clustering error is the number of points clustered incorrectly according to the generating random processes. Various clustering algorithms are evaluated based on process variance and the key issue of the rate at which algorithmic performance improves with increasing numbers of experimental replications. The model means can be selected by hand to test the separability of expected types of biological expression patterns. Alternatively, the model can be seeded by real data to test the expected precision of that output or the extent of improvement in precision that replication could provide. In the latter case, a clustering algorithm is used to form clusters, and the model is seeded with the means and variances of these clusters. Other algorithms are then tested relative to the seeding algorithm. Results are averaged over various seeds. Output includes error tables and graphs, confusion matrices, principal-component plots, and validation measures. Five algorithms are studied in detail: K-means, fuzzy C-means, self-organizing maps, hierarchical Euclidean-distance-based and correlation-based clustering. The toolbox is applied to gene-expression clustering based on cDNA microarrays using real data. Expression profile graphics are generated and error analysis is displayed within the context of these profile graphics. A large amount of generated output is available over the web.