Nature Precedings
Not a member yet
3446 research outputs found
Sort by
Statistical analysis of network data and evolution on GPUs: High-performance statistical computing
Get PDFNetwork analysis typically involves as set of repetitive tasks that are particularly amenable to poor-man's parallelization. This is therefore an ideal application are for GPU architectures, which help to alleviate the tedium inherent to statistically sound analysis of network data. Here we will illustrate the use of GPUs in a range of applications, which include percolation processes on networks, the evolution of protein-protein interaction networks, and the fusion of different types of biomedical and disease data in the context of molecular interaction networks. We will pay particular attention to the numerical performance of different routines that are frequently invoked in network analysis problems. We conclude with a review over recent developments in the generation of random numbers that address the specific requirements posed by GPUs and high-performance computing needs
Deletion of the trpc4 gene and its role in simple and complex strategic learning
Get PDFThe TRPC4 ion channel is expressed extensively in corticolimbic and a subpopulation of midbrain dopamine neurons. While TRPC4 knockout (KO) rats exhibit reduced sociability and social exploration, little is known about the role of TRPC4 in motivation and learning. To identify a function for TRPC4 channels in learning processes  we tested TRPC4 KO and normal wild type (WT) rats. TRPC4 KO and WT rats exhibited no differences in Y-­maze learning or simple discrimination learning. Furthermore, on a more complex serial reversal shift task designed  to assess strategic learning where the reward and non-­reward cues were repeatedly reversed between training sessions both TRPC4 KO and WT rats   performed equally well. Finally, we found no   performance differences when using a conditional reversal shift task where a tone signals the reversal of reward and non-reward cues within sessions. These data suggest that although TRPC4 channels may play a role in social interaction/anxiety  they exert a minimal role in simple and complex strategic learning
Integrative Approach - New Level Knowledge of Functions: Opportunities and Prospects
Get PDFIn this article, the example of the mechanisms of heart rhythmogenesis in the intact organism is used to demonstrate the new capabilities provided by an integrative approach. It is shown that the rhythm is formed in the brain, transmitted to the heart in the form of signals along the vagus nerves and reproduces the heart. Evidence: the heart rhythm reproduces the natural efferent signals in the vagus nerves in the cardio-respiratory synchronism and in the intact organism sino-atrial node performs the functions of the latent pacemaker. Integration of the two hierarchical levels of rhythmogenesis (brain and intracardiac) provides the reliability and functional perfection of cardiac rhythm generation in the body. It is expedient to extend the presented methodology for scientific analysis to other organism systems
Integration of the scientific literature into the Semantic Web: Facts from biomedical data resources
Get PDFScientific literature is the primary resource for relevant and innovative information. The integration of the literature with other data resources in the biomedical research community generates overhead that can be avoided through the used of Semantic Web Technology generating openly accessible data. Projects such as SESL and CALBC have producted significant amount of data that that are ready for exploitation.
The tutorial will teach different approaches on how to integrate the scientific literature with the content from biomedical databases, and will discuss the inferences that can be achieved. Furthermore, the tutorial will point to the resources that are ready for use and enable integration of the literature at your discretion. A good understanding of Semantic Web technology, ontologies, OWL and the existing biomedical data resources is advantageous to easily follow the tutorial
The computational magic of the ventral stream
Get PDFI argue that the sample complexity of (biological, feedforward) object recognition is mostly due to geometric image transformations and conjecture that a main goal of the ventral stream – V1, V2, V4 and IT – is to learn-and-discount image transformations.

In the first part of the paper I describe a class of simple and biologically plausible memory-based modules that learn transformations from unsupervised visual experience. The main theorems show that these modules provide (for every object) a signature which is invariant to local affine transformations and approximately invariant for other transformations. I also prove that,
in a broad class of hierarchical architectures, signatures remain invariant from layer to layer. The identification of these memory-based modules with complex (and simple) cells in visual areas leads to a theory of invariant recognition for the ventral stream.

In the second part, I outline a theory about hierarchical architectures that can learn invariance to transformations. I show that the memory complexity of learning affine transformations is drastically reduced in a hierarchical architecture that factorizes transformations in terms of the subgroup of translations and the subgroups of rotations and scalings. I then show how translations are automatically selected as the only learnable transformations during development by enforcing small apertures – eg small receptive fields – in the first layer.

In a third part I show that the transformations represented in each area can be optimized in terms of storage and robustness, as a consequence determining the tuning of the neurons in the area, rather independently (under normal conditions) of the statistics of natural images. I describe a model of learning that can be proved to have this property, linking in an elegant way the spectral properties of the signatures with the tuning of receptive fields in different areas. A surprising implication of these theoretical results is that the computational goals and some of the tuning properties of cells in the ventral stream may follow from symmetry properties (in the sense of physics) of the visual world through a process of unsupervised correlational learning, based on Hebbian synapses. In particular, simple and complex cells do not directly care about oriented bars: their tuning is a side effect of their role in translation invariance. Across the whole ventral stream the preferred features reported for neurons in different areas are only a symptom of the invariances computed and represented.

The results of each of the three parts stand on their own independently of each other. Together this theory-in-fieri makes several broad predictions, some of which are:

-invariance to small transformations in early areas (eg translations in V1) may underly stability of visual perception (suggested by Stu Geman);

-each cell’s tuning properties are shaped by visual experience of image transformations during developmental and adult plasticity;

-simple cells are likely to be the same population as complex cells, arising from different convergence of the Hebbian learning rule. The input to complex “complex” cells are dendritic branches with simple cell properties;

-class-specific transformations are learned and represented at the top of the ventral stream hierarchy; thus class-specific modules such as faces, places and possibly body areas should exist in IT;

-the type of transformations that are learned from visual experience depend on the size of the receptive fields and thus on the area (layer in the models) – assuming that the size increases with layers;

-the mix of transformations learned in each area influences the tuning properties of the cells oriented bars in V1+V2, radial and spiral patterns in V4 up to class specific tuning in AIT (eg face tuned cells);

-features must be discriminative and invariant: invariance to transformations is the primary determinant of the tuning of cortical neurons rather than statistics of natural images.

The theory is broadly consistent with the current version of HMAX. It explains it and extend it in terms of unsupervised learning, a broader class of transformation invariance and higher level modules. The goal of this paper is to sketch a comprehensive theory with little regard for mathematical niceties. If the theory turns out to be useful there will be scope for deep mathematics, ranging from group representation tools to wavelet theory to dynamics of learning
Determination of Soybean Oil, Protein and Amino Acid Residues in Soybean Seeds by High Resolution Nuclear Magnetic Resonance (NMRS) and Near Infrared (NIRS)
Get PDFA detailed account is presented of our high resolution nuclear magnetic resonance (HR-NMR) and near infrared (NIR) calibration models, methodologies and validation procedures, together with a large number of composition analyses for soybean seeds. NIR calibrations were developed based on both HR-NMR and analytical chemistry reference data for oil and twelve amino acid residues in mature soybeans and soybean embryos. This is our first report of HR-NMR determinations of amino acid profiles of proteins from whole soybean seeds, without protein extraction from the seed. It was found that the best results for both oil and protein calibrations were obtained with a Partial Least Squares Regression (PLS-1) analysis of our extensive NIR spectral data, acquired with either a DA7000 Dual Diode Array (Si and InGaAs detectors) instrument or with several Fourier Transform NIR (FT-NIR) spectrometers equipped with an integrating sphere/InGaAs detector accessory. In order to extend the bulk soybean samples calibration models to the analysis of single soybean seeds, we have analized in detail the component NIR spectra of all major soybean constituents through spectral deconvolutions for bulk, single and powdered soybean seeds. Baseline variations and light scattering effects in the NIR spectra were corrected, respectively, by calculating the first-order derivatives of the spectra and the Multiplicative Scattering Correction (MSC). The single soybean seed NIR spectra are broadly similar to those of bulk whole soybeans, with the exception of minor peaks in single soybean NIR spectra in the region from 950 to 1,000 nm. Based on previous experience with bulk soybean NIR calibrations, the PLS-1 calibration model was selected for protein, oil and moisture calibrations that we developed for single soybean seed analysis. In order to improve the reliability and robustness of our calibrations with the PLS-1 model we employed standard samples with a wide range of soybean constituent compositions: from 34% to 55% for protein, from 11% to 22% for oil and from 2% to 16% for moisture. Such calibrations are characterized by low standard errors and high degrees of correlation for all major soybean constituents. Morever, we obtained highly resolved NIR chemical images for selected regions of mature soybean embryos that allow for the quantitation of oil and protein components. Recent developments in high-resolution FT-NIR microspectroscopy extend the NIR sensitivity range to the picogram level, with submicron spatial resolution in the component distribution throughout intact soybean seeds and embryos. Such developments are potentially important for biotechnology applications that require rapid and ultra- sensitive analyses, such as those concerned with high-content microarrays in Genomics and Proteomics research. Other important applications of FT-NIR microspectroscopy are envisaged in biomedical research aimed at cancer prevention, the early detection of tumors by NIR-fluorescence, and identification of single cancer cells, or single virus particles in vivo by super-resolution microscopy/ microspectroscopy
A Systems Biology Approach towards Deciphering the Unfolded Protein Response in Huntington's Disease
Get PDFAlthough the disease causing gene huntingtin has been known for some time, the exact cause of neuronal cell death during _Huntington's disease_ (HD) remains unknown. One potential mechanism contributing to the massive loss of neurons in HD brains might be the _Unfolded Protein Response_ (UPR) which is activated by accumulation of misfolded proteins in the endoplasmic reticulum (ER). As an adaptive response, UPR upregulates transcription of chaperones, temporarily attenuating new translation and activates protein degradation via the proteasome. However, at high levels of ER stress, UPR signalling can contribute to neuronal apoptosis.

Our primary aims include (a) construction of the UPR signalling network, (b) curation and bioinformatical identification of UPR target genes and finally (c) examination of HD gene expression data sets for UPR transcriptional signatures and differential regulation of UPR pathways.

The UPR signalling pathway is reconstructed based on literature review and using the "Unified Interactome database":http://www.unihi.org. Lists of UPR target genes detected by previous experiments or as predicted by computational analysis are compiled. This allows us to perform enrichment analysis for differential HD gene expression and to assess whether UPR expression signatures are prominent during HD pathogenesis.

Results: The canonical UPR pathway is complemented with additional protein interaction data allowing us to assess its embedding into the cellular context and to identify potential modifiers as well as novel drug targets.

Conclusions: The in depth systems biology analysis can give us valuable insights about the involvement of the UPR in HD.

Music can reduce cognitive dissonance
Get PDFThe fundamental cognitive functions of music in the brain have not been known and evolutionary reasons for musical abilities seem mysterious. A recent hypothesis suggested that a fundamental function of music has been to help mitigating cognitive dissonances. A cognitive dissonance is "a discomfort caused by holding conflicting cognitions" simultaneously; it usually leads to devaluation of conflicting knowledge. Since every concept implies some degree of contradictions to other knowledge, unmitigated cognitive dissonances could prevent evolution of cognition. Thus music might be fundamental for the evolution of cognition. Here we provide experimental confirmation of this hypothesis using a classical paradigm known to induce a cognitive dissonance and devaluation of a dissonant object; in presence of music devaluation has not occurred
Integration of Co-expression Networks for Gene Clustering
Get PDFSimultaneous overexpression or underexpression of multiple genes, used in various forms as probes in the high-throughput microarray experiments, facilitates the identification of their underlying functional proximity. This kind of functional associativity (or conversely the separability) between the genes can be represented proficiently using co-expression networks. The extensive repository of diversified microarray data encounters a recent problem of multi-experimental data integration for the aforesaid purpose. This paper highlights a novel integration method of gene co-expression networks, based on the search for their consensus network, derived from diverse microarray experimental data for the purpose of clustering. The proposed methodology avoids the bias arising from missing value estimation. The method has been applied on microarray datasets arising from different category of experiments to integrate them. The consensus network, thus produced, reflects robustness based on biological validation
Genomic Replikin Count Predicts Increased Lethality of Malaria
Get PDFGenomic Replikin Counts predict both the increase and the decrease of lethality of malari