SEARCH · PubMed Health

Results for “Network inference”

Explore indexed PubMed citations for clinical trials, systematic reviews and public health research. Read source abstracts and follow each citation to its original PubMed record.

Quote a phrase for an exact phrase match. Source license links do not imply unrestricted reuse.

At least 667 records · Page 37Linked to original sources

PubMed1986

Evidence for the persistence of a decondensed chromosome scaffold in the interphase nucleus.

Nuclei of in vitro cultured bovine liver cells, deprived of the membranes by Triton X-100, were treated with 2 M-NaCl and DNase. Changes in ultrastructure and protein composition were studied at successive steps during treatment. Electron micrographs of nuclei treated with 2 M-NaCl showed a peripheral lamina and an internal system of randomly coiled filaments embedded in a mass of DNA fibres. After partial removal of the DNA the filaments could be seen to serve as backbones for the DNA attachment. Artificial redistribution occurring during fixation with glutaraldehyde suggests that the salt-resistant filaments are not stably cross-bridged into a three-dimensional network. The existence of reversible cross-bridges in vivo cannot be excluded, however. From the available data it is inferred that the filaments represent a decondensed from of the chromosome scaffolds and play a basic role in the organization of the genome throughout the nuclear cycle.

Animals↗

PubMed2000

The primary structure of the subunit in Bacillus thermoamyloliquefaciens KP1071 molecular weight 540,000 homohexameric alpha-glucosidase II belonging to the glycosyl hydrolase family 31.

The gene that coded for the subunit of an molecular weight (Mr) 540,000 homohexameric alpha-glucosidase II (alpha-D-glucoside glucohydrolase, EC 3.2.1.20) produced by Bacillus thermoamyloliquefaciens KP1071 (FERM-P8477) growing at 30 to 66 degrees C was expressed in Escherichia coli HB101. The resulting homohexameric enzyme had a half-life of 10 min at 80 degrees C. Its purification and characterization showed that the enzyme was identical with the native one except for the latter deleting 7 N-terminal residues found in the former. The primary sequence of the subunit with 787 residues and an Mr of 91,070 deduced from the gene was 24-34% identical to the corresponding sequences of 15 alpha-glucosidases in the glycosyl hydrolase family 31 from 14 eukaryotic origins and the archaeon Sulfolobus solfataricus 98/2. From the sequence analysis by the neural network method of Rost and Sander [Rost, B. and Sander, C., Proteins: Struct. Funct. Genet., 19, 55-72 (1994)], we inferred that alpha-glucosidase II might make each subunit of 3 secondary structural regions, i.e., one N-terminal beta region, one central alpha/beta region with two catalytic residues Asp407 and Asp484, and one C-terminal beta region.

Amino Acid Sequence↗

PubMed2004

Design and diversity in bacterial chemotaxis: a comparative study in Escherichia coli and Bacillus subtilis.

Comparable processes in different species often involve homologous genes. One question is whether the network structure, in particular the feedback control structure, is also conserved. The bacterial chemotaxis pathways in E. coli and B. subtilis both regulate the same task, namely, excitation and adaptation to environmental signals. Both pathways employ many orthologous genes. Yet how these orthologs contribute to network function in each organism is different. To investigate this problem, we propose what is to our knowledge the first computational model for B. subtilis chemotaxis and compare it to previously published models for chemotaxis in E. coli. The models reveal that the core control strategy for signal processing is the same in both organisms, though in B. subtilis there are two additional feedback loops that provide an additional layer of regulation and robustness. Furthermore, the network structures are different despite the similarity of the proteins in each organism. These results demonstrate the limitations of pathway inferences based solely on homology and suggest that the control strategy is an evolutionarily conserved property.

Acclimatization↗

PubMed1999

A unifying review of linear gaussian models.

Factor analysis, principal component analysis, mixtures of gaussian clusters, vector quantization, Kalman filter models, and hidden Markov models can all be unified as variations of unsupervised learning under a single basic generative model. This is achieved by collecting together disparate observations and derivations made by many previous authors and introducing a new way of linking discrete and continuous state models using a simple nonlinearity. Through the use of other nonlinearities, we show how independent component analysis is also a variation of the same basic generative model. We show that factor analysis and mixtures of gaussians can be implemented in autoencoder neural networks and learned using squared error plus the same regularization term. We introduce a new model for static data, known as sensible principal component analysis, as well as a novel concept of spatially adaptive observation noise. We also review some of the literature involving global and local mixtures of the basic models and provide pseudocode for inference and learning for all the basic models.

Algorithms↗

PubMed2000

Direct evidence for local oscillatory current sources and intracortical phase gradients in turtle visual cortex.

Visual stimuli induce oscillations in the membrane potential of neurons in cortices of several species. In turtle, these oscillations take the form of linear and circular traveling waves. Such waves may be a consequence of a pacemaker that emits periodic pulses of excitation that propagate across a network of excitable neuronal tissue or may result from continuous and possibly reconfigurable phase shifts along a network with multiple weakly coupled neuronal oscillators. As a means to resolve the origin of wave propagation in turtle visual cortex, we performed simultaneous measurements of the local field potential at a series of depths throughout this cortex. Measurements along a single radial penetration revealed the presence of broadband current sources, with a center frequency near 20 Hz (gamma band), that were activated by visual stimulation. The spectral coherence between sources at two well-separated loci along a rostral-caudal axis revealed the presence of systematic timing differences between localized cortical oscillators. These multiple oscillating current sources and their timing differences in a tangential plane are interpreted as the neuronal activity that underlies the wave motion revealed in previous imaging studies. The present data provide direct evidence for the inference from imaging of bidirectional wave motion that the stimulus-induced electrical waves in turtle visual cortex correspond to phase shifts in a network of coupled neuronal oscillators.

Animals↗

PubMed1998

Long-range connections synchronize rather than spread intrathalamic oscillations: computational modeling and in vitro electrophysiology.

A thalamic network model was developed based on recent data regarding heterogeneous thalamic reticular (RE) cell axonal arborizations that indicate at least two projection patterns, short-range cluster projections and long-range diffuse projections. The model was constrained based on expected convergence and the biophysical properties of RE and thalamocortical (TC) cells and their synapses. The model reproduced in vitro synchronous slow (3-Hz) oscillatory activity and the known effects of T-channel blockade and cholecystokinin (CCK) application on this activity. Whereas previous models used the speed at which approximately 3-Hz oscillations propagate in vitro to infer the spatial extent of intrathalamic projections, we found that, so long as the gamma-aminobutyric acid-B synaptic conductance was adjusted appropriately, a network with only short-range projections and another network with both short- and long-range projections could both produce physiologically realistic propagation speeds. Although the approximately 3-Hz oscillations propagated at similar speeds in both networks, phase differences between oscillatory activity at different locations in the network were much smaller in the network containing both short- and long-range projections. We measured phase differences in vitro and found that they were similar to those that arise in the network containing both short- and long-range projections but are inconsistent with the much larger phase differences that occur in the network containing only short-range projections. These results suggest that, although they extend much further than do short-range cluster projections, long-range diffuse projections do not spread activity over greater distances or increase the speed at which intrathalamic oscillations propagate. Instead, diffuse projections may function to synchronize activity and minimize phase shifts across thalamic networks. One prediction of this hypothesis is that, immediately after a collision between propagating oscillations, phase gradients should vary smoothly across the thalamic slice. The model also predicts that phase shifts between oscillatory activity at different points along a thalamic slice should be unaffected by T-channel blockers and decreased by suppression of synaptic transmission or application of CCK.

Animals↗

PubMed2006

Similarity classifier using similarity measure derived from Yu's norms in classification of medical data sets.

A new approach using a similarity measure based on Yu's norms is presented for the detection of erythemato-squamous diseases, diabetes, breast cancer, lung cancer and lymphography. The domain contains records of patients with known diagnoses. The results are very promising with all data sets and (in conclusion, can be drawn that) a similarity model derived from Yu's norms could be used for the diagnosis of patients taking into consideration the error rate. A similarity classifier derived from Yu's norms was used to detect the six erythemato-squamous diseases when 34 features defining six disease indications were used as inputs. The results confirmed that the proposed model has potential in detecting the erythemato-squamous diseases. The similarity model derived from Yu's norms achieved an accuracy rate (97.8%) which was higher than that of the stand-alone neural network model or the ANFIS model suggested in Ubeyli and Güler [Automatic detection of erythemato-squamous diseases using adaptive neuro-fuzzy inference systems, Comput. Biol. Med. 35 (2005) 421-433] or the similarity model based on Łukasiewicz similarity [Luukka and Leppälampi, Similarity classifier with generalized mean applied to medical data, Comput. Biol. Med. 36 (2006) 1026-1040]. With PIMA Indian diabetes, the detection model has an error rate of about 24% which is much better than the overall rate of 33% for diabetes. Also, a classifier was applied to the lung cancer data set and the results were to my knowledge better than before. When the lung cancer data were preprocessed with an entropy minimization technique and the classifier with similarity based on Yu's norm was applied, 99.99% accuracy was achieved. The use of this preprocessing method enhanced the results over 30%. In lymphography, entropy minimization also enhanced the results remarkably and 86.2% accuracy was achieved.

Breast Neoplasms↗

PubMed1992

Object-oriented fuzzy expert system for on-line diagnosing and control of bioprocesses.

An object-oriented fuzzy expert system to support on-line control of an automated fermentation plant is described. The major elements of the system consist of a fuzzy inference engine, a database, a knowledge base, and an expression evaluater. The expression evaluater calculates specific rates for growth, and substrate and product formation at different physiological states during the cultivation from the measured data. The specific rates are then compared with the standard target rates stored in the database. If differences outside the set tolerances were observed, the inference engine analyses the reasons for the faults on the basis of the knowledge represented in the form of a knowledge network and fuzzy membership functions of the process variables. The fuzzy expert system was developed on the basis of a shell constructed by using the object oriented Smalltalk/V Mac programming environment, with Lac-tobacillus casei lactic acid fermentation as the example of process application.

Biotechnology↗

PubMed2005

Reconstructing pre-acidification pH for an acidified Scottish loch: a comparison of palaeolimnological and modelling approaches.

We reconstruct the pre-acidification pH of the Round Loch of Glenhead for 1800 AD using three diatom-pH transfer functions and a diatom-cladocera modern analogue technique (MAT), and compare these palaeo-data with hindcast data for the loch using the dynamic catchment acidification model MAGIC. We assess the accuracy of the transfer functions by comparing pH inferences from contemporary sediment and sediment trap diatom samples from the lake with measured pH from the UK Acid Waters Monitoring Network. The results from the transfer functions estimate the pH in 1800 to have been between 5.5. and 5.7, the MAT approach estimates pH at 5.8 and the MAGIC hindcast (for 1850) is pH 6.1. Whilst we have no independent method of assessing which of these values is most accurate, the disagreement between the two approaches indicates that further work is needed to resolve the discrepancies.

Acid Rain↗

PubMed2001

Crystal structures of the family 9 carbohydrate-binding module from Thermotoga maritima xylanase 10A in native and ligand-bound forms.

The C-terminal module of the thermostable Thermotoga maritima xylanase 10A (CBM9-2) is a family 9 carbohydrate-binding module that binds to amorphous and crystalline cellulose and a range of soluble di- and monosaccharides as well as to cello and xylo oligomers of different degrees of polymerization [Boraston, A. B., Creagh, A. L., Alam, Md. M., Kormos, J. M., Tomme, P., Haynes, C. A., Warren, R. A. J., and Kilburn, D. G. (2001) Biochemistry 40, 6240-6247]. The crystal structure of CBM9-2 has been determined by the multiwavelength anomalous dispersion method to 1.9 A resolution. CBM9-2 assumes a beta-sandwich fold and contains three metal binding sites. The bound metal atoms, which are most likely calcium cations, are in an octahedral coordination. The crystal structures of CBM9-2 in complex with glucose and cellobiose were also determined in order to identify the sugar-binding site and provide insight into the structural basis for sugar binding by CBM9-2. The sugar-binding site is a solvent-exposed slot sufficient in depth, width, and length to accommodate a disaccharide. Two tryptophan residues are stacked together on the surface of the protein forming the sugar-binding site. From the complex structures with glucose and cellobiose, it was inferred that CBM9-2 binds exclusively to the reducing end of mono-, di-, and oligosaccharides with an intricate hydrogen-bonding network involving mainly charged residues, as well as stacking interactions by Trp175 and Trp71. The binding interactions are limited to disaccharides as was expected from calorimetric data. Comparison of the glucose and cellobiose complexes revealed surprising differences in binding of these two substrates by CBM9-2. Cellobiose was found to bind in a distinct orientation from glucose, while still maintaining optimal stacking and electrostatic interactions with the reducing end sugar.

Amino Acid Sequence↗

PubMed2003

Visualization and interpretation of protein networks in Mycobacterium tuberculosis based on hierarchical clustering of genome-wide functional linkage maps.

Genome-wide functional linkages among proteins in cellular complexes and metabolic pathways can be inferred from high throughput experimentation, such as DNA microarrays, or from bioinformatic analyses. Here we describe a method for the visualization and interpretation of genome-wide functional linkages inferred by the Rosetta Stone, Phylogenetic Profile, Operon and Conserved Gene Neighbor computational methods. This method involves the construction of a genome-wide functional linkage map, where each significant functional linkage between a pair of proteins is displayed on a two-dimensional scatter-plot, organized according to the order of genes along the chromosome. Subsequent hierarchical clustering of the map reveals clusters of genes with similar functional linkage profiles and facilitates the inference of protein function and the discovery of functionally linked gene clusters throughout the genome. We illustrate this method by applying it to the genome of the pathogenic bacterium Mycobacterium tuberculosis, assigning cellular functions to previously uncharacterized proteins involved in cell wall biosynthesis, signal transduction, chaperone activity, energy metabolism and polysaccharide biosynthesis.

Bacterial Proteins↗

PubMed2005

Molecular pathogenesis of follicular lymphoma: a cross talk of genetic and immunologic factors.

Follicular lymphoma (FL) is the second most frequent type of non-Hodgkin's lymphoma in adults. The disease is characterized by an indolent course with frequent relapses. Ultimately, resistance to chemotherapy or transformation to a more aggressive phase of the disease in the form of diffuse large B-cell lymphoma develops, and patients die as a result of their disease. Median survival is 8 to 10 years. The range is very wide, however, with patients surviving for more than 15 years and 10% to 15% of the patients who run a rapidly fatal course and die within 3 years after diagnosis. The translocation t(14;18) is the basic molecular defect in FL and results in protection from apoptosis by aberrant overexpression of bcl-2 protein. Accumulation of genomic alterations and clonal selection account for subsequent progression and transformation. Recently, the role of the immunologic microenvironment of FL in determining clinical behavior and prognosis has been substantiated. Combined genetic and immunologic data may now support a model for the development of FL as a disease of functional B cells in which specific molecular alterations infer intrinsic growth properties of the tumor cells as well as dictate a specific functional cross talk with the immunologic regulatory network resulting in extrinsic growth support. These insights may lead to improvement of risk stratification, but most importantly will provide tools for developing new targets and strategies for treatment.

Adult↗

PubMed2004

A recurrent connectionist model of person impression formation.

Major findings in impression formation are reviewed and modeled from a connectionist perspective. The findings are in the areas of primacy and recency in impression formation, asymmetric diagnosticity of ability- and morality-related traits, increased recall for trait-inconsistent information, assimilation and contrast in priming, and discounting of trait inferences by situational information. The majority of these phenomena are illustrated with well-known experiments and simulated with an autoassociative network architecture with linear activation update and using the delta learning algorithm for adjusting the connection weights. All of the simulations successfully reproduced the empirical findings. Moreover, the proposed model is shown to be consistent with earlier algebraic models of impression formation (Anderson, 1981; Busemeyer, 1991; Hogarth and Einhorn, 1992). The discussion centers on how our model compares to other connectionist approaches to impression formation and how it may contribute to a more parsimonious and unified theory of person perception.

Brain↗

PubMed1997

Cardiovascular regulation: a modelling approach.

A detailed analysis of autonomic cardiovascular control (ACVC) may provide a key to a better understanding of the mechanisms underlying postflight orthostatic hypotension. The central substrate of human ACVC is not directly accessible to measurements and observation in space research. Modelling--supporting inference and physiological reasoning--is a valuable tool to disclose its involvement We are currently determining the suitability of artificial neural networks (ANN's) as a model of the central substrate of ACVC. Having conducted a number of experiments with simulated tilt test data to clarify the choice of input coding and of architectural biases in network training we will now report on the approximation of data obtained from human subjects during preparation of the German MIR'97 and D-2 missions.

Adaptation, Physiological↗

PubMed2025

Gene regulatory network structure informs the distribution of perturbation effects.

Gene regulatory networks (GRNs) govern many core developmental and biological processes underlying human complex traits. Even with broad-scale efforts to characterize the effects of molecular perturbations and interpret gene coexpression, it remains challenging to infer the architecture of gene regulation in a precise and efficient manner. Key properties of GRNs, like hierarchical structure, modular organization, and sparsity, provide both challenges and opportunities for this objective. Here, we seek to better understand properties of GRNs using a new approach to simulate their structure and model their function. We produce realistic network structures with a novel generating algorithm based on insights from small-world network theory, and we model gene expression regulation using stochastic differential equations formulated to accommodate modeling molecular perturbations. With these tools, we systematically describe the effects of gene knockouts within and across GRNs, finding a subset of networks that recapitulate features of a recent genome-scale perturbation study. With deeper analysis of these exemplar networks, we consider future avenues to map the architecture of gene expression regulation using data from cells in perturbed and unperturbed states, finding that while perturbation data are critical to discover specific regulatory interactions, data from unperturbed cells may be sufficient to reveal regulatory programs.

Gene Regulatory Networks↗

PubMedDate not supplied

A neural network approach to approximating MAP in belief networks.

Bayesian belief networks (BBN) are a widely studied graphical model for representing uncertainty and probabilistic interdependence among variables. One of the factors that restricts the model's wide acceptance in practical applications is that the general inference with BBN is NP-hard. This is also true for the maximum a posteriori probability (MAP) problem, which is to find the most probable joint value assignment to all uninstantiated variables, given instantiation of some variables in a BBN. To circumvent the difficulty caused by MAP's computational complexity, we suggest in this paper a neural network approximation approach. With this approach, a BBN is treated as a neural network without any change or transformation of the network structure, and the node activation functions are derived based on an energy function defined over a given BBN. Three methods are developed. They are the hill-climbing style discrete method, the simulated annealing method, and the continuous method based on the mean field theory. All three methods are for BBN of general structures, with the restriction that nodes of BBN are binary variables. In addition, rules for applying these methods to noisy-or networks are also developed, which may lead to more efficient computation in some cases. These methods' convergence is analyzed, and their validity tested through a series of computer experiments with two BBN of moderate size and complexity. Although additional theoretical and empirical work is needed, the analysis and experiments suggest that this approach may lead to effective and accurate approximation for MAP problems.

Algorithms↗

PubMed2001

Two-dimensional fluorometry coupled with artificial neural networks: a novel method for on-line monitoring of complex biological processes.

The use of two-dimensional scanning fluorometry as an on-line, noninvasive, in situ bioreactor monitoring technique is extended to complex bioprocesses using mixed cultures, with particular attention to biofilm systems. Using the example of spectra subtraction, it is demonstrated that established methods for fluorescence data analysis have a limited capability of utilizing overall fluorometric information. Artificial neural networks (ANNs) are introduced as a novel nonlinear and nonmechanistic technique for interpreting the highly complex fluorescence maps. It is shown that ANNs are able to infer process performance parameters in a pattern recognition approach, based on the entire fluorescence "fingerprint" of the biological system. The studies were carried out using an extractive membrane bioreactor (EMB) for the degradation of chlorinated organic compounds, operating with mixed cultures. Model pollutants em- ployed were 1,2-dichloroethane, 3-chloro-4-methylaniline, and p-toluidine.

Biodegradation, Environmental↗

PubMed1999

Collective motions in HIV-1 reverse transcriptase: examination of flexibility and enzyme function.

In order to study the inferences of structure for mechanism, the collective motions of the retroviral reverse transcriptase HIV-1 RT (RT) are examined using the Gaussian network model (GNM) of proteins. This model is particularly suitable for elucidating the global dynamic characteristics of large proteins such as the presently investigated heterodimeric RT comprising a total of 982 residues. Local packing density and coordination order of amino acid residues is inspected by the GNM to determine the type and range of motions, both at the residue level and on a global scale, such as the correlated movements of entire subdomains. Of the two subunits, p66 and p51, forming the RT, only p66 has a DNA-binding cleft and a functional polymerase active site. This difference in the structure of the two subunits is shown here to be reflected in their dynamic characteristics: only p66 has the potential to undergo large-scale cooperative motions in the heterodimer, while p51 is essentially rigid. Taken together, the global motion of the RT heterodimer is comprised of movements of the p66 thumb subdomain perpendicular to those of the p66 fingers, accompanied by anticorrelated fluctuations of the RNase H domain and p51 thumb, thus providing information about the details of one processivity mechanism. A few clusters of residues, generally distant in sequence but close in space, are identified in the p66 palm and connection subdomains, which form the hinge-bending regions that control the highly concerted motion of the subdomains. These regions include the catalytically active site and the non-nucleoside inhibitor binding pocket of p66 polymerase, as well as sites whose mutations have been shown to impair enzyme activity. It is easily conceivable that this hinge region, indicated by GNM analysis to play a critical role in modulating the global motion, is locked into an inactive conformation upon binding of an inhibitor. Comparative analysis of the dynamic characteristics of the unliganded and liganded dimers indicates severe repression of the mobility of the p66 thumb in RT's global mode, upon binding of non-nucleoside inhibitors.

Binding Sites↗