Decoding semi-constrained brain activity from FMRI using support vector machines and gaussian processes

Predicting a particular cognitive state from a specific pattern of fMRI voxel values is still a methodological challenge. Decoding brain activity is usually performed in highly controlled experimental paradigms characterized by a series of distinct states induced by a temporally constrained experimental design. In more realistic conditions, the number, sequence and duration of mental states are unpredictably generated by the individual, resulting in complex and imbalanced fMRI data sets. This study tests the classification of brain activity, acquired on 16 volunteers using fMRI, during mental imagery, a condition in which the number and duration of mental events were not externally imposed but self-generated. To deal with these issues, two classification techniques were considered (Support Vector Machines, SVM, and Gaussian Processes, GP), as well as different feature extraction methods (General Linear Model, GLM and SVM). These techniques were combined in order to identify the procedures leading to the highest accuracy measures. Our results showed that 12 data sets out of 16 could be significantly modeled by either SVM or GP. Model accuracies tended to be related to the degree of imbalance between classes and to task performance of the volunteers. We also conclude that the GP technique tends to be more robust than SVM to model unbalanced data sets.     

Constructing large-scale genetic maps using an evolutionary strategy algorithm

This article is devoted to the problem of ordering in linkage groups with many dozens or even hundreds of markers. The ordering problem belongs to the field of discrete optimization on a set of all possible orders, amounting to n!/2 for n loci; hence it is considered an NP-hard problem. Several authors attempted to employ the methods developed in the well-known traveling salesman problem (TSP) for multilocus ordering, using the assumption that for a set of linked loci the true order will be the one that minimizes the total length of the linkage group. A novel, fast, and reliable algorithm developed for the TSP and based on evolution-strategy discrete optimization was applied in this study for multilocus ordering on the basis of pairwise recombination frequencies. The quality of derived maps under various complications (dominant vs. codominant markers, marker misclassification, negative and positive interference, and missing data) was analyzed using simulated data with approximately 50-400 markers. High performance of the employed algorithm allows systematic treatment of the problem of verification of the obtained multilocus orders on the basis of computing-intensive bootstrap and/or jackknife approaches for detecting and removing questionable marker scores, thereby stabilizing the resulting maps. Parallel calculation technology can easily be adopted for further acceleration of the proposed algorithm. Real data analysis (on maize chromosome 1 with 230 markers) is provided to illustrate the proposed methodology.     

Handicapped individua in evolutionary processes

Usually, when considering an evolutionary development, the fittest individua and their succesfull mutations recive most attention. The less fit ones, existing due to selection-mutation equilibria, are regarded, as an inevitable waste. It appears that these handicaped individua play an important role in evolution, as they make escapes from evolutionary traps possible. It is concluded that crucial improvements come from mutations of mutants. The implications seem to account to some extent for the observed speed and mode of evolution. An appropriate stochastic model is introduced, numerical experiments reported and results discussed.     

Considering evolutionary processes in the use of single-locus genetic data for conservation,with examples from the Lepidoptera

The increasing popularity of molecular taxonomy will undoubtedly have a major impact on the practice of conservation biology. The appeal of such approaches is undeniable since they will clearly be an asset in rapid biological assessments of poorly known taxa or unexplored areas, and for discovery of cryptic biodiversity. However, as an approach for diagnosing units for conservation, some caution is warranted. The essential issue is that mitochondrial DNA variation is unlikely to be causally related to, and thus correlated with, ecologically important components of fitness. This is true for DNA barcoding, molecular taxonomy in general, or any technique that relies on variation at a single, presumed neutral locus. Given that natural selection operates on a time scale that is often much more rapid than the rates of mutation and allele frequency changes due to genetic drift, neutral genetic variation at a single locus can be a poor predictor of adaptive variation within or among species. Furthermore, reticulate processes, such as introgressive hybridization, may also constrain the utility of molecular taxonomy to accurately detect significant units for conservation. A survey of published genetic data from the Lepidoptera indicates that these problems may be more prevalent than previously suspected. Molecular approaches must be used with caution for conservation genetics which is best accomplished using large sample sizes over extensive geography in addition to data from multiple loci. Matthew L. Forister, Chris C. Nice and James A. Fordyce contributed equally to this paper.     

Prediction in evolutionary systems

Despite its explanatory clout, the theory of evolution has thus far compiled a modest record with respect to predictive power—that other major hallmark of scientific theories. This is considered by many to be an acceptable limitation of a theory that deals with events and processes that are intrinsically random (and historic). However, whether this is an inherent restriction or simply the sign of an incomplete theory is an open question. In an attempt to help answer that question, we propose a classification scheme for several types of prediction that might occur with regard to evolutionary systems, then explore the nature of these predictions in a system that simulates the evolution of neural architectures. This provides a platform from which to consider the relevance of such observations for real biological systems and illuminates a variety of key issues pertaining to prediction in those environments.     

An investigation of the evolutionary origin of reciprocal communication using simulated autonomous agents

How does communication originates in a population of originally non-communicating individuals? Providing an answer to this question from a neo-Darwinian epistemological perspective is not a trivial task. The reason is that, for non-communicating agents, the capabilities of emitting signals and responding to them are both adaptively neutral traits if they are not simultaneously present. Research studies based on rather general and theoretically oriented evolutionary simulation models have, so far, demonstrated that at least two different processes can account for the origin of communication. On the one hand, communicative behaviour may first evolve in a non-communicative context and only subsequently acquire its adaptive function. On the other hand, communication may originate thanks to cognitive constraints; that is, communication may originate thanks to the existence of neural substrates that are common to the signalling and categorising capabilities. This article provides a proof-of-concept demonstration of the origin of communication in a novel-simulated scenario in which groups of two homogeneous (i.e. genetically identical) agents exploit reciprocal communication to develop common perceptual categories and to perform a collective task. In particular, in circumstances in which communication is evolutionarily advantageous, simulated agents evolve from scratch social behaviour through acoustic interactions. We look into the phylogeny of successful communication protocol, and we describe the evolutionary phenomena that, in early evolutionary stages, paved the way for the subsequent development of reciprocal communication, categorisation capabilities and successful cooperative strategies.     

Considering evolutionary processes in conservation biology

Conservation biologists assign population distinctiveness by classifying populations as evolutionarily significant units (ESUs). Historically, this classification has included ecological and genetic data. However, recent ESU concepts, coupled with increasing availability of data on neutral genetic variation, have led to criteria based exclusively on molecular phylogenies. We argue that the earlier definitions of ESUs, which incorporated ecological data and genetic variation of adaptive significance, are more relevant for conservation. Furthermore, this dichotomous summary (ESU or not) of a continuum of population differentiation is not adequate for determining appropriate management actions. We argue for a broader categorization of population distinctiveness based on concepts of ecological and genetic exchangeability (sensu Templeton).     

Fast and robust characterization of time-heterogeneous sequence evolutionary processes using substitution mapping

Genes and genomes do not evolve similarly in all branches of the tree of life. Detecting and characterizing the heterogeneity in time, and between lineages, of the nucleotide (or amino acid) substitution process is an important goal of current molecular evolutionary research. This task is typically achieved through the use of non-homogeneous models of sequence evolution, which being highly parametrized and computationally-demanding are not appropriate for large-scale analyses. Here we investigate an alternative methodological option based on probabilistic substitution mapping. The idea is to first reconstruct the substitutional history of each site of an alignment under a homogeneous model of sequence evolution, then to characterize variations in the substitution process across lineages based on substitution counts. Using simulated and published datasets, we demonstrate that probabilistic substitution mapping is robust in that it typically provides accurate reconstruction of sequence ancestry even when the true process is heterogeneous, but a homogeneous model is adopted. Consequently, we show that the new approach is essentially as efficient as and extremely faster than (up to 25 000 times) existing methods, thus paving the way for a systematic survey of substitution process heterogeneity across genes and lineages.     

Regulation of proteolytic processes using antisense RNA genes htpR and lon of Escherichia coli in hetero- and homologous genetic systems

Possibility of correction of proteolytic processes in cells of Escherichia coli and Pseudomonas aeruginosa has been studied. For this purpose recombinant plasmids directing the synthesis of antisense RNAs were constructed. In Ps. aeruginosa the synthesis of htpR antisense RNA resulted in 2.5-fold reduction of the intensity of degradation of 3H-puromycin polypeptides under heat shock conditions. An antisense RNA complementary to the 5'- end of E. coli lon gene decreased the same index to the level observed in lon- mutants. Genes homologous to htpR and lon genes of E. coli were found in Pseudomonas: bacteria in hybridisation experiments. This finding suggests that the genetic system of heat shock in these microorganisms is organized in a similar manner.     

Predicting the insurgence of human genetic diseases associated to single point protein mutations with support vector machines and evolutionary information

MOTIVATION: Human single nucleotide polymorphisms (SNPs) are the most frequent type of genetic variation in human population. One of the most important goals of SNP projects is to understand which human genotype variations are related to Mendelian and complex diseases. Great interest is focused on non-synonymous coding SNPs (nsSNPs) that are responsible of protein single point mutation. nsSNPs can be neutral or disease associated. It is known that the mutation of only one residue in a protein sequence can be related to a number of pathological conditions of dramatic social impact such as Alzheimer's, Parkinson's and Creutzfeldt-Jakob's diseases. The quality and completeness of presently available SNPs databases allows the application of machine learning techniques to predict the insurgence of human diseases due to single point protein mutation starting from the protein sequence. RESULTS: In this paper, we develop a method based on support vector machines (SVMs) that starting from the protein sequence information can predict whether a new phenotype derived from a nsSNP can be related to a genetic disease in humans. Using a dataset of 21 185 single point mutations, 61% of which are disease-related, out of 3587 proteins, we show that our predictor can reach more than 74% accuracy in the specific task of predicting whether a single point mutation can be disease related or not. Our method, although based on less information, outperforms other web-available predictors implementing different approaches. AVAILABILITY: A beta version of the web tool is available at http://gpcr.biocomp.unibo.it/cgi/predictors/PhD-SNP/PhD-SNP.cgi     

Modeling of biological processes using self-cycling fermentation and genetic algorithms

Self-cycling fermentation (SCF) was coupled with a genetic algorithm (GA) to provide a simple system for evaluating biological models. The SCF provided the necessary system excitation and data "richness" required to completely define the fitted biological models. The solution scheme based on the GA avoided the computational difficulties often associated with calculus-based nonlinear regression techniques, resulting in rapid and accurate convergence. After validating the mathematical approach, data from the SCF obtained under denitrifying conditions were fitted successfully to an established model using the GA. Finally, data obtained in the SCF for the removal of phenol were used to compare multiple models. This work suggests that the SCF, in conjunction with the GA, provides a coherent system that can facilitate the characterization of biological systems.     

Evolution of a genetic code simulated with the computer

A simple selforganizing model system of molecules is considered and it is demonstrated by a computer simulation, that a genetic code of 16 elements (aminoacids) can gradually be formed by such a system in the course of many generations. By a number of rare chance events, each suppressing other events of equal a priori probability, a single code results out of an immense number of possible codes of the same a priori probability. The result is discussed in relation to the uniqueness of the genetic code in living systems. The computer simulation emphasizes a particular step in a model pathway discussed elsewhere consisting of many assumed physicochemical steps leading to a genetic apparatus.     

Possible evolutionary steps in the genetic code

In mammalian mitochondrial codes, fourfold codons wobble-pair with UNN anticodons so that U wobbles with U, C, A and G. Twofold pyrimidine-terminated codons pair with GNN and twofold purine-terminated codons pair with UNN. These properties enable a prediction to be made for evolution of the universal genetic code. It was postulated (1) that an archetypal code of 16 quartets coded for 15 amino acids. If this code used UNN anticodons, then duplication of tRNA genes, followed by mutations in the anticodons and aminoacylation sites, would give rise to the present universal code.     

Variation in evolutionary processes at different codon positions

Evolutionary studies commonly model single nucleotide substitutions and assume that they occur as independent draws from a unique probability distribution across the sequence studied. This assumption is violated for protein-coding sequences, and we consider modeling approaches where codon positions (CPs) are treated as separate categories of sites because within each category the assumption is more reasonable. Such "codon-position" models have been shown to explain the evolution of codon data better than homogenous models in previous studies. This paper examines the ways in which codon-position models outperform homogeneous models and characterizes the differences in estimates of model parameters across CPs. Using the PANDIT database of multiple species DNA sequence alignments, we quantify the differences in the evolutionary processes at the 3 CPs in a systematic and comprehensive manner, characterizing previously undescribed features of protein evolution. We relate our findings to the functional constraints imposed by the genetic code, protein function, and the types of mutation that cause synonymous and nonsynonymous codon changes. The results increase our understanding of selective constraints and could be incorporated into phylogenetic analyses or gene-finding techniques in the future. The methods used are extended to an overlapping reading frame data set, and we discover that overlapping reading frames do not necessarily cause more stringent evolutionary constraints.     

Oscillating systems with cointegrated phase processes

Journal of Mathematical Biology - We present cointegration analysis as a method to infer the network structure of a linearly phase coupled oscillating system. By defining a class of oscillating...