Pietro Di Lena

Deep architectures for protein contact map prediction

MOTIVATION Residue-residue contact prediction is important for protein structure prediction and other applications. However, the accuracy of current contact predictors often barely exceeds 20% on long-range contacts, falling short of the level required for ab initio structure prediction. RESULTS Here, we develop a novel machine learning approach for contact map prediction using three steps of increasing resolution. First, we use 2D recursive neural networks to predict coarse contacts and orientations between secondary structure elements. Second, we use an energy-based method to align secondary structure elements and predict contact probabilities between residues in contacting alpha-helices or strands. Third, we use a deep neural network architecture to organize and progressively refine the prediction of contacts, integrating information over both space and time. We train the architecture on a large set of non-redundant proteins and test it on a large set of non-homologous domains, as well as on the set of protein domains used for contact prediction in the two most recent CASP8 and CASP9 experiments. For long-range contacts, the accuracy of the new CMAPpro predictor is close to 30%, a significant increase over existing approaches. AVAILABILITY CMAPpro is available as part of the SCRATCH suite at http://scratch.proteomics.ics.uci.edu/. CONTACT pfbaldi@uci.edu SUPPLEMENTARY INFORMATION Supplementary data are available at Bioinformatics online.

FT-COMAR

UNLABELLED Fault Tolerant Contact Map Reconstruction (FT-COMAR) is a heuristic algorithm for the reconstruction of the protein three-dimensional structure from (possibly) incomplete (i.e. containing unknown entries) and noisy contact maps. FT-COMAR runs within minutes, allowing its application to a large-scale number of predictions. AVAILABILITY http://bioinformatics.cs.unibo.it/FT-COMAR

Blurring contact maps of thousands of proteins: what we can learn by reconstructing 3D structure

BackgroundThe present knowledge of protein structures at atomic level derives from some 60,000 molecules. Yet the exponential ever growing set of hypothetical protein sequences comprises some 10 million chains and this makes the problem of protein structure prediction one of the challenging goals of bioinformatics. In this context, the protein representation with contact maps is an intermediate step of fold recognition and constitutes the input of contact map predictors. However contact map representations require fast and reliable methods to reconstruct the specific folding of the protein backbone.MethodsIn this paper, by adopting a GRID technology, our algorithm for 3D reconstruction FT-COMAR is benchmarked on a huge set of non redundant proteins (1716) taking random noise into consideration and this makes our computation the largest ever performed for the task at hand.ResultsWe can observe the effects of introducing random noise on 3D reconstruction and derive some considerations useful for future implementations. The dimension of the protein set allows also statistical considerations after grouping per SCOP structural classes.ConclusionsAll together our data indicate that the quality of 3D reconstruction is unaffected by deleting up to an average 75% of the real contacts while only few percentage of randomly generated contacts in place of non-contacts are sufficient to hamper 3D reconstruction.

Genome-Wide Identification of Bcl11b Gene Targets Reveals Role in Brain-Derived Neurotrophic Factor Signaling

B-cell leukemia/lymphoma 11B (Bcl11b) is a transcription factor showing predominant expression in the striatum. To date, there are no known gene targets of Bcl11b in the nervous system. Here, we define targets for Bcl11b in striatal cells by performing chromatin immunoprecipitation followed by high-throughput sequencing (ChIP-seq) in combination with genome-wide expression profiling. Transcriptome-wide analysis revealed that 694 genes were significantly altered in striatal cells over-expressing Bcl11b, including genes showing striatal-enriched expression similar to Bcl11b. ChIP-seq analysis demonstrated that Bcl11b bound a mixture of coding and non-coding sequences that were within 10 kb of the transcription start site of an annotated gene. Integrating all ChIP-seq hits with the microarray expression data, 248 direct targets of Bcl11b were identified. Functional analysis on the integrated gene target list identified several zinc-finger encoding genes as Bcl11b targets, and further revealed a significant association of Bcl11b to brain-derived neurotrophic factor/neurotrophin signaling. Analysis of ChIP-seq binding regions revealed significant consensus DNA binding motifs for Bcl11b. These data implicate Bcl11b as a novel regulator of the BDNF signaling pathway, which is disrupted in many neurological disorders. Specific targeting of the Bcl11b-DNA interaction could represent a novel therapeutic approach to lowering BDNF signaling specifically in striatal cells.

Fast overlapping of protein contact maps by alignment of eigenvectors

MOTIVATION Searching for structural similarity is a key issue of protein functional annotation. The maximum contact map overlap (CMO) is one of the possible measures of protein structure similarity. Exact and approximate methods known to optimize the CMO are computationally expensive and this hampers their applicability to large-scale comparison of protein structures. RESULTS In this article, we describe a heuristic algorithm (Al-Eigen) for finding a solution to the CMO problem. Our approach relies on the approximation of contact maps by eigendecomposition. We obtain good overlaps of two contact maps by computing the optimal global alignment of few principal eigenvectors. Our algorithm is simple, fast and its running time is independent of the amount of contacts in the map. Experimental testing indicates that the algorithm is comparable to exact CMO methods in terms of the overlap quality, to structural alignment methods in terms of structure similarity detection and it is fast enough to be suited for large-scale comparison of protein structures. Furthermore, our preliminary tests indicates that it is quite robust to noise, which makes it suitable for structural similarity detection also for noisy and incomplete contact maps. AVAILABILITY Available at http://bioinformatics.cs.unibo.it/Al-Eigen.

Decidable properties for regular cellular automata

We investigate decidable properties for regular cellular automata. In particular, we show that regularity itself is an undecidable property and that nilpotency, equicontinuity and positively expansiveness became decidable if we restrict to regular cellular automata.

Optimal global alignment of signals by maximization of Pearson correlation

The problem of detecting the similarity between noisy signals obtained from electronic instrument measurements arises in several different contexts and it is approached with specific strategies accordingly. In this paper we propose a simple and general method for the comparison of noisy signals of different lengths. Assuming any a-priori knowledge about two noisy signals, their degree of similarity can be detected by computing the global alignment that maximizes their Pearson correlation. The Pearson correlation coefficient is a widely used measure of linear dependence between two random variables of the same length. The optimal alignment of two signals with respect to the Pearson correlation identifies the sub-regions of the two signals that exhibit the highest pairwise degree of similarity. We show that the optimal alignment of two signals by maximization of the Pearson correlation can be computed in (quadratic) polynomial-time by a simple application of the Needleman-Wunsch algorithm. Our approach can be used for the comparison of one-dimensional signals, multi-dimensional signals and multiple-alignments of (one-dimensional or multi-dimensional) signals.

Reconstruction of 3D structures from protein contact maps

Proteins are large organic compounds made of amino acids arranged in a linear chain (primary structure). Most proteins fold into unique three-dimensional (3D) structures called interchangeably tertiary, folded, or native structures. Discovering the tertiary structure of a protein (Protein Folding Problem) can provide important clues about how the protein performs its function and it is one of the most important problems in Bioinformatics. A contact map of a given protein P is a binary matrix M such that Mi,j = 1 iff the physical distance between amino acids i and j in the native structure is less than or equal to a pre-assigned threshold t. The contact map of each protein is a distinctive signature of its folded structure. Predicting the tertiary structure of a protein directly from its primary structure is a very complex and still unsolved problem. An alternative and probably more feasible approach is to predict the contact map of a protein from its primary structure and then to compute the tertiary structure starting from the predicted contact map. This last problem has been recently proven to be NP-Hard [6]. In this paper we give a heuristic method that is able to reconstruct in a few seconds a 3D model that exactly matches the target contact map. We wish to emphasize that our method computes an exact model for the protein independently of the contact map threshold. To our knowledge, our method outperforms all other techniques in the literature [5,10,17,19] both for the quality of the provided solutions and for the running times. Our experimental results are obtained on a non-redundant data set consisting of 1760 proteins which is by far the largest benchmark set used so far. Average running times range from 3 to 15 seconds depending on the contact map threshold and on the size of the protein. Repeated applications of our method (starting from randomly chosen distinct initial solutions) show that the same contact map may admit (depending on the threshold) quite different 3D models. Extensive experimental results show that contact map thresholds ranging from 10 to 18 Angstrom allow to reconstruct 3D models that are very similar to the proteins native structure. Our Heuristic is freely available for testing on the web at the following url: http://vassura.web.cs.unibo.it/cmap23d/

MIMO: an efficient tool for molecular interaction maps overlap

BackgroundMolecular pathways represent an ensemble of interactions occurring among molecules within the cell and between cells. The identification of similarities between molecular pathways across organisms and functions has a critical role in understanding complex biological processes. For the inference of such novel information, the comparison of molecular pathways requires to account for imperfect matches (flexibility) and to efficiently handle complex network topologies. To date, these characteristics are only partially available in tools designed to compare molecular interaction maps.ResultsOur approach MIMO (Molecular Interaction Maps Overlap) addresses the first problem by allowing the introduction of gaps and mismatches between query and template pathways and permits -when necessary- supervised queries incorporating a priori biological information. It then addresses the second issue by relying directly on the rich graph topology described in the Systems Biology Markup Language (SBML) standard, and uses multidigraphs to efficiently handle multiple queries on biological graph databases. The algorithm has been here successfully used to highlight the contact point between various human pathways in the Reactome database.ConclusionsMIMO offers a flexible and efficient graph-matching tool for comparing complex biological pathways.

Periodic Orbits and Dynamical Complexity in Cellular Automata

We investigate the relationships between dynamical complexity and the set of periodic configurations of surjective Cellular Automata. We focus on the set of strictly temporally periodic configurations, i.e., the set of those configurations which are temporally but not spatially periodic for a given surjective automaton. The cardinality of this set turns out to be inversely related to the dynamical complexity of the cellular automaton. In particular, we show that for surjective Cellular Automata, the set of strictly temporally periodic configurations has strictly positive measure if and only if the cellular automaton is equicontinuous. Furthermore, we show that the set of strictly temporally periodic configurations is dense for almost equicontinuous surjective cellular automata, while it is empty for the positively expansive ones. In the class of additive cellular automata, the set of strictly temporally periodic points can be either dense or empty. The latter happens if and only if the cellular automaton is topologically transitive. This is not true for general transitive Cellular Automata, where the set of of strictly temporally periodic points can be non-empty and non-dense.