Search PubMed⌕ Search

SEARCH · Search PubMed

Results for “Parallel Algorithms”

Search indexed PubMed citations on genomics, clinical trials, systematic reviews and public health. Explore titles, authors and supplied subject terms, then open the PubMed record.

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

At least 217 records · Page 12Linked to original sources
PubMed1992

Grand challenges in biomedical computing.

Advances in computing technology (both algorithms and hardware) over the next several years promise to make increasingly sophisticated computer modeling of biomedical phenomena a routine part of biomedical research. Improvements in both the absolute speed of processors and in their programming and graphics interfaces will allow nonexpert users to bring computing power equivalent to the supercomputers of a few years ago to bear on routine research problems and to display complex data in understandable ways (visualization). Although biomedical applications have traditionally not driven the leading edge of computing and supercomputing, such applications are increasingly being ported to advanced parallel and vector processors. This paper summarizes the current state of biomedical computing, citing examples of the best practice in research today. A number of projects enabled by advanced computing from various subdisciplines are described. Trends in technology for both inexpensive (workstation) and high-end computing (vector supercomputers and parallel processors) are cited; the implications of these for biomedical computing are discussed. "Grand challenges" in biomedical computing, i.e., computational problems of major scientific importance that are beyond our current capabilities but that might be achieved in a 5-year time frame, are outlined.

Attitude to Computers↗
PubMed1990

An optimal control model for maximum-height human jumping.

To understand how intermuscular control, inertial interactions among body segments, and musculotendon dynamics coordinate human movement, we have chosen to study maximum-height jumping. Because this activity presents a relatively unambiguous performance criterion, it fits well into the framework of optimal control theory. The human body is modeled as a four-segment, planar, articulated linkage, with adjacent links joined together by frictionless revolutes. Driving the skeletal system are eight musculotendon actuators, each muscle modeled as a three-element, lumped-parameter entity, in series with tendon. Tendon is assumed to be elastic, and its properties are defined by a stress-strain curve. The mechanical behavior of muscle is described by a Hill-type contractile element, including both series and parallel elasticity. Driving the musculotendon model is a first-order representation of excitation-contraction (activation) dynamics. The optimal control problem is to maximize the height reached by the center of mass of the body subject to body-segmental, musculotendon, and activation dynamics, a zero vertical ground reaction force at lift-off, and constraints which limit the magnitude of the incoming neural control signals to lie between zero (no excitation) and one (full excitation). A computational solution to this problem was found on the basis of a Mayne-Polak dynamic optimization algorithm. Qualitative comparisons between the predictions of the model and previously reported experimental findings indicate that the model reproduces the major features of a maximum-height squat jump (i.e. limb-segmental angular displacements, vertical and horizontal ground reaction forces, sequence of muscular activity, overall jump height, and final lift-off time).

Biomechanical Phenomena↗
PubMed1981

Analysis and predication of structural motifs in the glycolytic enzymes.

Protein crystallography has determined the three-dimensional structures of 10 of the 13 enzymes of the glycolytic pathway. Diagrams and details of these enzyme structures are given in the paper. Most of the enzyme domains are variations and extensions of a many (4--9)-stranded, predominantly or totally parallel, beta-sheet that is shielded from solvent by alpha-helices (i.e. alpha/beta structures). There are strong structural similarities between the domains of some, but not all, of the enzymes. In particular the dinucleotide binding fold of lactate dehydrogenase and the beta-barrel of triose phosphate isomerase are found in other domains. General rules governing the topology and packing of alpha-helices against a beta-sheet provide a basis for the combinatorial prediction of the tertiary fold of glycolytic domains from their amino acid sequence and observed secondary structure. The predication algorithm demonstrates that there are severe restrictions on the number of possible structures. However, these restrictions do not fully explain some of the remarkable structural similarities between different enzymes that probably result from evolution from a common ancestor.

Animals↗
PubMed1991

A distributed-parameter model of the myelinated nerve fiber.

This paper presents a new model for the characterization of electrical activity in the nodal, paranodal and internodal regions of isolated amphibian and mammalian myelinated nerve fibers. It differs from previous models in the following ways: (1) in its ability to incorporate detailed anatomical and electrophysiological data; (2) in its approach to the myelinated nerve fiber as a multi-axial cable; and (3) in the numerical algorithm used to obtain distributed model equation solutions for potential and current. The morphometric properties are taken from detailed electron microscopic anatomical studies (Berthold & Rydmark, 1983a, Experientia 39, 964-976). The internodal axolemma is characterized as an excitable membrane and model-generated nodal and internodal membrane action potentials are presented. A system of describing equations for the equivalent network model is derived, based on the application of Kirchoff's Current Law, which take the form of multiple cross-coupled parabolic partial differential equations. An implicit numerical integration method is developed and the numerical solution implemented on a parallel processor. Non-uniform spatial step sizes are used, enabling detailed representation of the nodal region while minimizing the number of total segments necessary to represent the overall fiber. Conduction velocities of 20.2 m sec-1 at 20 degrees C for a 15 microns diameter amphibian fiber and 57.6 m sec-1 at 37 degrees C for a 17.5 microns diameter mammalian fiber are achieved, which agrees qualitatively with published experimental data at similar temperatures (Huxley & Stämpfli, 1949, J. Physiol., Lond. 108, 315-339; Rasminsky, 1973, Arch, Neurol. 28, 287-292). The simulation results demonstrate the ability of this model to produce detailed representations of the transaxonal, transmyelin and transfiber potentials and currents, as well as the longitudinal extra-axonal, periaxonal and intra-axonal currents. Also indicated is the potential contribution of the paranodal axolemma to nodal activity as well as the presence of significant longitudinal currents in the periaxonal space adjacent to the node of Ranvier.

Action Potentials↗
PubMed1988

Binding of the bovine basic pancreatic trypsin inhibitor (Kunitz) to human alpha-, beta- and gamma-thrombin; a kinetic and thermodynamic study.

Kinetic and thermodynamic parameters for the binding of the bovine basic pancreatic trypsin inhibitor (BPTI, Kunitz inhibitor) to human alpha-, beta- and gamma-thrombin have been determined, between 5 and 45 degrees C, at pH 7.5. BPTI-binding properties to human thrombins have been analyzed in parallel with those of serine (pro)enzymes acting on cationic and non-cationic substrates, with particular reference to the bovine beta-trypsin/BPTI system. The observed binding behaviour of BPTI to human alpha-, beta- and gamma-thrombin has been related to the inferred stereochemistry of the enzyme/inhibitor contact region(s).

Algorithms↗
PubMed1991

Parallel cascade identification and kernel estimation for nonlinear systems.

We consider the representation and identification of nonlinear systems through the use of parallel cascades of alternating dynamic linear and static nonlinear elements. Building on the work of Palm and others, we show that any discrete-time finite-memory nonlinear system having a finite-order Volterra series representation can be exactly represented by a finite number of parallel LN cascade paths. Each LN path consists of a dynamic linear system followed by a static nonlinearity (which can be a polynomial). In particular, we provide an upper bound for the number of parallel LN paths required to represent exactly a discrete-time finite-memory Volterra functional of a given order. Next, we show how to obtain a parallel cascade representation of a nonlinear system from a single input-output record. The input is not required to be Gaussian or white, nor to have special autocorrelation properties. Next, our parallel cascade identification is applied to measure accurately the kernels of nonlinear systems (even those with lengthy memory), and to discover the significant terms to include in a nonlinear difference equation model for a system. In addition, the kernel estimation is used as a means of studying individual signals to distinguish deterministic from random behaviour, in an alternative to the use of chaotic dynamics. Finally, an alternate kernel estimation scheme is presented.

Algorithms↗
PubMed1992

Detection, classification and 3D reconstruction of biological macromolecules on hypercube computers.

In this work we present results of the mapping on hypercube computers of some of the key steps involved in the procedure for 3D structural determination from transmission electron microscopy images. The goal is the introduction of parallel processing tools in the field of electron microscopy image processing. We show how the rich topology of the hypercube, combined with an efficient programming strategy, allows for order-of-magnitude increase in computational capacity for such time-consuming tasks as calculation of multidimensional FFT's, cross-correlation coefficients, fuzzy partitioning functionals and the filtered back-projection 3D reconstruction method.

Algorithms↗
PubMed1992

Insect motion perception.

The first step in this work of reconstruction of a theory of insect vision was to demonstrate that visual behaviour relies on scanning by self-motion and apparently involves measurement of angular velocities of contrasts moving across the eye. The next step was to demonstrate that parallax is also significant as a way of segmenting the visual scene into separate objects. There followed a series of experiments to rule out the existing theory that motion perception depends on autocorrelation, and at the same time an alternative theory was developed. The new theory assumes that at the level of the optic medulla there are numerous parallel channels on each visual axis, representing different neurons, all looking out for their specific combination of signals. The combinations are formed by positive, negative or no-change temporal contrasts at two adjacent visual axes at two successive times, forming 3(4) = 81 possible templates. Simulation of this highly parallel system shows that it can represent the moving image in a compact form that would be adequate to explain what is known for motion and form vision (but not colour vision) in insects. Form, like colour, would be seen as the ratio of numbers of responses of particular templates, in the same way that colours are seen as ratios of responses of receptors for different wavelengths.(ABSTRACT TRUNCATED AT 250 WORDS)

Algorithms↗
PubMed1989

Assessment of the vasomotility of epicardial coronary arteries with quantitative coronary angiography.

Quantitative in vivo analysis of the vasomotility of epicardial coronary arteries is based on the measurement of changes of the vessel diameters. Vessel contours can be determined with the help of a precision caliper by the investigator or with computer-assisted geometrical analysis systems applying a contour detection algorithm; the variability of the results from repeated coronary diameter measurements in the identical film frame is comparably low with all systems (less than or equal to 0.12 mm standard deviation). When investigating the influence of an intervention on coronary vasomotility the variability of the measurements can only be kept low by careful standardization of the entire method. Repeated coronary angiograms are performed in identical angiographic projections (mono- or biplane) with standardization of the inspiratory status of the patient and of the rate of contrast material injection (automatic injection pump). For quantitative analysis all coronary segments with a diameter greater than 1 mm which are clearly outlined, free from overlaps, and mainly run parallel to the image plane, are selected by the investigator in preferably end-diastolic cineframes. Whereas with a caliper and with most of the semi-automatic edge detection systems the segment diameter can only be measured at particular sites defined by the investigator, few systems are able to analyze and average the diameter over the entire segment length. The variability of measurements of the minimal diameters of coronary stenoses in different cine frames is particularly high; therefore, only distinct changes of this parameter (e.g. with the CAAS-system greater than 0.24 mm) may be considered significant.(ABSTRACT TRUNCATED AT 250 WORDS)

Cineangiography↗
PubMed2026

FFC: a scalable FASTA compressor.

SUMMARY: FASTA is a widely used text-based format for storing nucleotide and protein sequences. The existing FASTA compressors usually focus on (slightly) improving the compression ratio, not on practical performance. We present FFC, a scalable FASTA compressor that achieves average compression speeds 4.7× and 11.4× higher than two high-performance compressors, zstd and NAF, respectively, across a benchmark set of seven single genomes. It also delivers average decompression speeds 3.5× and 2.7× higher than zstd and NAF, respectively. Although a chunk-based zstd variant with parallel decompression, pzstd, almost matches FFC speed, its compression ratio is on average by 23% worse than FFC's. For the experiment, a 14-core workstation and a RAM disk (to reduce the impact of I/O) were used. AVAILABILITY AND IMPLEMENTATION: FFC is freely available at github.com/kowallus/ffc and also as a Zenodo repository at 10.5281/zenodo.18892353, and the used datasets at 10.5281/zenodo.18873744.

Data Compression↗
PubMed1990

Calculation of site affinity constants and cooperativity coefficients for binding of ligands and/or protons to macromolecules. I. Generation of partition functions and mass balance equations.

The thermodynamics of binding of a ligand A and/or proton H to a macromolecule M is treated by the partition function method. In complex systems, the representation of the equilibria by means of cumulative constants beta PQR used as coefficients in partition functions ZM, ZA, and ZH is ill-suited to least-squares refinement procedures because the cumulative constants are interrelated by common cooperativity functions gamma j(i) and common site affinity constants kappa j. There is therefore the need to express ZM, ZA, ZH as functions of site constants kappa j and cooperativity coefficients bj. This is done by developing an algebra of partition functions based on the following concepts: (i) factorability of partition functions; (ii) binary generating function Jj = (1 + kappa j[Y])i tau for each class j of sites, represented by column (Jj) and row (Jj) vectors; (iii) cooperativity between sites of one class described by functions gamma j(i), represented by diagonal matrices gamma j; (iv) probability of finding microspecies represented by elements of tensor product matrix Ll = (J1)[J2]; (v) statistical factors mij obtained from Newton polynomials, Jj; (vi) power operators Oi', O(i-l)', and O(i tau-l)', transforming vectors Jj; and (vii) operators Oi or O(i-l) indicating tensor products of i or (i-l) vectors Jj. Vectors Jj combined in tensors Ll give rise to both an affinity/cooperativity space and a parallel index space. The partition functions ZM, ZA, and ZH and the total amounts TM, TA, and TH can be obtained as an appropriate sum of elements of matrices Ll, each of which is represented in an index space by a combination p1, p2,...q1, q2,...r1, r2,... of indices ij. From these indices the contribution of that element to partition function ZM, ZA, or ZH and to total amount TM, TA, or TH is calculated in the affinity/cooperativity space as product of factors: [i tau !/i !(i tau-i)!]kappa ij(exp[bj (i-1)i])[X]i, i being any index p, q, r and X any component M, A, or H. Future applications of this algorithm to practical problems of macromolecule-ligand-proton equilibria are outlined.

Binding Sites↗
PubMed1992

Molecular modeling of the amphipathic helices of the plasma apolipoproteins.

In this paper we propose a classification of the amphipathic helical repeats occurring in the plasma apolipoprotein sequences. It is based upon the calculation of the molecular hydrophobicity potential around the helical segments. The repeats were identified using a new autocorrelation matrix, based upon similarities of hydrophobic and hydrophilic properties of the amino acid residues within the apolipoprotein sequences. The helices were constructed by molecular modeling, the molecular hydrophobicity potential was calculated, and isopotential contour lines drawn around the helices yielded a three-dimensional visualization of the hydrophobicity potential. Two classes of apolipoproteins could be differentiated by comparing the hydrophobic angles obtained by projection of the isopotential contour lines on a plane perpendicular to the long axis of the helix. The isopotential contour lines around apo AI, AIV, and E are more hydrophilic than hydrophobic, whereas they are of similar intensity for apo AII, CI, and CIII. In both cases discoidal lipid-protein complexes are generated, with the amphipathic helices around the edge of the lipid core. The long axis of the helices is oriented parallel to the phospholipid acyl chains and the hydrophilic side of the helix toward the aqueous phase. As a result of the differences in hydrophobicity potential, the contact between the hydrophobic side of the helices and the phospholipid acyl chains is larger for apo AII, CI, and CIII than for the other apolipoproteins. This might account for the greater stability of the discoidal complexes generated between phospholipids and these apoproteins.

Algorithms↗
PubMed1992

A new motion responsive pacemaker: first clinical experience with an acceleration sensor pacemaker.

A new accelerometer-based adaptive rate pacemaker (EXCEL VR) was evaluated to investigate its behavior at nominal settings during treadmill exercise testing and postural changes. Eight patients with sinus rhythm were selected to compare intrinsic heart rate to sensor mediated rate. Throughout exercise treadmill testing, changes in the sensor mediated rate closely paralleled actual physiological changes. The Pearson product moment correlation of pacing rate with sinus frequency, calculated for all patients, was r = 0.82 (P = 0.001). Change in the patient's physical position resulted in immediate change in sensor rate, which corresponded appropriately to the type of position change and activity level. Average (+/- SD) pacing rate was 62.4 +/- 2.7 beats/min supine; 67 +/- 3.8 beats/min sitting; 69.8 +/- 6.4 beats/min standing; 81.6 +/- 8.7 beats/min slow walking; and 96.8 +/- 1.3 beats/min fast walking. After 4 minutes of recovery, the average pacing rate dropped to 65 +/- 3 beats/min. The interaction between the accelerometer and the pulse generator at nominal settings was accurate and infrequently required the use of its many programming options. The accelerometer sensor and pulse generator algorithm in this device during postural change and exercise resulted in physiological-like changes in sensor mediated heart rate.

Acceleration↗
PubMed1987

Two-dimensional 1H NMR study of human ubiquitin: a main chain directed assignment and structure analysis.

The 1H resonances of human ubiquitin were studied by two-dimensional nuclear magnetic resonance techniques. A recently introduced assignment algorithm termed the main chain directed (MCD) assignment [Englander, S. W., & Wand, A. J. (1987) Biochemistry 26, 5953-5958] was applied. This approach relies on an ordered series of searches for prescribed patterns of connectivities in two-dimensional J-correlated and nuclear Overhauser effect spectra and centers on the dipolar interactions involving main-chain amide NH, alpha-CH, and beta-CH. Unlike the sequential assignment procedure, the MCD approach does not rest upon definition of side-chain J-coupled networks and is generally not sequential with the primary sequence of the protein. The various MCD patterns and the general algorithm are reiterated and applied to the analysis of human ubiquitin. With this algorithm, the vast majority of amino acid residue amide NH-C alpha H-C beta H J-coupled subspin systems could be associated with and aligned within units of secondary structure without any knowledge of the identity of the side chains. This greatly simplified recognition of side-chain spin systems by restricting their identity. Essentially complete resonance assignments are presented. The MCD method is compared with the sequential assignment method in some detail. The MCD method is highly amenable to automation. Human ubiquitin is found, at pH 5.8 and 30 degrees C, to be composed of an extensive beta-sheet structure involving five strands. Three of these strands form an antiparallel set sharing a common strand and have a parallel orientation to two antiparallel strands. Two helical segments were also observed. The largest, spanning 13 residues, shows dipolar interactions consistent with an alpha-helix while the smaller 4-residue helical segment appears, on the basis of observed nuclear Overhauser effects, to be a 3(10) helix. Five classical tight turns could be demonstrated.

Humans↗
PubMed2026

Atlas-level single-cell integration and clustering-free differential expression analysis with GEDI 2.0.

MOTIVATION: GEDI is a generative framework for multi-sample, multi-condition single-cell analysis that performs batch correction, latent representation learning, and clustering-free differential expression within a unified model. However, the original implementation suffered from prohibitive memory use and runtime, preventing its application to modern atlas-scale datasets. RESULTS: We present GEDI 2.0, a complete high-performance reimplementation featuring a standalone C++ computational core with pre-allocated workspaces, strict sparse-matrix preservation, optimized BLAS routines, and multi-threaded block-coordinate descent. Across extensive benchmarks spanning up to 500 000 cells and 10 000 features, GEDI 2.0 achieves 40%-63.6% mean reduction in peak memory, 2.98× mean single-threaded speedups, and up to 11.5× acceleration with parallel execution, while maintaining full numerical equivalence to the original method. These improvements enable GEDI 2.0 to analyze million-cell datasets, a scale not achievable with the legacy implementation. GEDI 2.0 provides R and Python interfaces and seamless interoperability with common single-cell workflows. AVAILABILITY AND IMPLEMENTATION: Source code, documentation, reproducible codebase, and tutorials are available at https://github.com/csglab/gedi2.

Single-Cell Analysis↗
PubMedDate not supplied

Appraising the instantaneous secretory rates of luteinizing hormone and testosterone in response to selective mu opiate receptor blockade in late pubertal boys.

The pulsatile properties of gonadotropin and testosterone release were examined before and after chronic mu opiate receptor blockade with naltrexone, 50 mg every other day, in four normal boys in late puberty (ages 14 8/12 to 15 1/12 years). The nature of spontaneous secretory events was appraised for immunoactive LH and testosterone in blood withdrawn every 20 minutes for 24 hours, using a novel, discrete deconvolution algorithm to estimate apparent instantaneous secretory rates. The application of this methodology revealed that the frequency of discrete LH instantaneous secretory rates increased after mu opiate receptor blockade (P = 0.011). More strikingly, all parameters of testosterone secretory events responded significantly to mu opiate receptor blockade, including increases in mean estimated secretory rate (+47%, P = 0.02), testosterone pulse frequency (+ 64%, P less than 0.001) and amplitude (+ 20%, P = 0.027). Correspondingly, decreases in testosterone interpulse secretory intervals (-35%, P = 0.001), secretory pulse duration (-19%, P = 0.042) and interpulse valley duration (-35%, P = 0.006) also were noted. There was a prominent diurnal rhythm in testosterone secretion with maximal values in the morning and late evening, and marked reductions in the afternoon, sometimes to prepubertal levels. This variation in the testosterone secretory profile paralleled that of LH. In response to naltrexone, the FSH concentration series showed a significant increase in the mean FSH concentration (+ 18%) P = 0.003) and mean peak amplitude (+ 15%, P = 0.002). These data provide indirect evidence of functional coupling of the opiate system with the hypothalamic GnRH pulse generator.(ABSTRACT TRUNCATED AT 250 WORDS)

Adolescent↗
PubMed1990

Distributed parallel processing in the vertical vestibulo-ocular reflex: learning networks compared to tensor theory.

The vestibulo-ocular reflex (VOR) is capable of producing compensatory eye movements in three dimensions. It utilizes the head rotational velocity signals from the semicircular canals to control the contractions of the extraocular muscles. Since canal and muscle coordinate frames are not orthogonal and differ from one another, a sensorimotor transformation must be produced by the VOR neural network. Tensor theory has been used to construct a linear transformation that can model the three-dimensional behavior of the VOR. But tensor theory does not take the distributed, redundant nature of the VOR neural network into account. It suggests that the neurons subserving the VOR, such as vestibular nucleus neurons, should have specific sensitivity-vectors. Actual data, however, are not in accord. Data from the cat show that the sensitivity-vectors of vestibular nucleus neurons, rather than aligning with any specific vectors, are dispersed widely. As an alternative to tensor theory, we modeled the vertical VOR as a three-layered neural network programmed using the back-propagation learning algorithm. Units in mature networks had divergent sensitivity-vectors which resembled those of actual vestibular nucleus neurons in the cat. This similarity suggests that the VOR sensorimotor transformation may be represented redundantly rather than uniquely. The results demonstrate how vestibular nucleus neurons can encode the VOR sensorimotor transformation in a distributed manner.

Animals↗
PubMed1992

Limitations of Augustinsson plots.

The Augustinsson plot transforms the Michaelis-Menten equation by graphing velocity, v, as a function of velocity/substrate, v/s. The plot is linear with a negative slope for a single class of saturable sites and curvilinear for multiple classes of sites. Statistically, Augustinsson plots suffer from the limitation that there is no true independent variable. Experimental errors form a Gaussian distribution that rests neither parallel to the v axis nor v/s axis, but rather along a line emanating from the origin. Widening the range of s values produces increasingly large errors in the calculated Km and Vm values, especially when multiple sites are present. Exact data containing no error produces accurate constants using either the Michaelis-Menten or Augustinsson equation. With Km and Vm constants previously derived by fitting data with the statistically valid Michaelis-Menten equation, the Augustinsson plot may be used accurately to display multiple site data.

Algorithms↗