SEARCH · PubMed Health

Results for “Deep generative models”

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 199 records · Page 11Linked to original sources

PubMed2003

Inhibition of purified factor Xa amidolytic activity may not be predictive of inhibition of in vivo thrombosis: implications for identification of therapeutically active inhibitors.

OBJECTIVE: In this study we test the hypothesis that blood/plasma-based prothrombinase assays, rather than inhibition of purified factor Xa (fXa), are predictive of in vivo antithrombotic activity. METHODS AND RESULTS: Six fXa inhibitors with equivalent nanomolar Ki were studied in thrombin generation assays using human plasma/blood and endogenous macromolecular substrate. In all assays, benzamidine inhibitors were more potent (100 to 800 nmol/L) than the aminoisoquinolines (5 to 58 micromol/L) or neutral inhibitors (3 to 10 micromol/L). A similar rank order of compound inhibition was also seen in purified prothrombinase assays as well as in a rabbit model of deep vein thrombosis. CONCLUSIONS: Assays using prothrombinase with protein substrates are better predictors of in vivo efficacy than fXa Ki using amidolytic substrates.

Animals↗

PubMed1984

Theoretical temperature distributions produced by an annular phased array-type system in CT-based patient models.

Theoretical calculations for the specific absorption rate (SAR) and the resulting temperature distributions produced by an annular phased array (APA)-type system are made. The finite element numerical method is used in the formulation of both the electromagnetic (EM) and thermal boundary value problems. A number of detailed two-dimensional patient models based on CT-scan data from the pelvic, visceral, and thoracic regions are generated to simulate a variety of tumor locations and surrounding normal tissues. The SAR values from the EM solution are put into the bioheat transfer equation, and steady-state temperature distributions are calculated for a wide range of blood flow rates. Based on our theoretical modeling, the APA shows no preferential heating of superficial over deep-seated tumors. However, in most cases for all three regions of the human trunk only fair thermal profiles (therapeutic area near 60%) are obtained in tumors with little or no blood flow and poor temperature patterns (therapeutic area less than 50%) are found in tumors with moderate to high perfusion rates. These theoretical calculations should aid the clinician in the evaluation of the effectiveness of APA-type devices in heating tumors located in the trunk region.

Humans↗

PubMed2026

Human Monocytic Models Reveal Genotype-Dependent Inflammatory Programs in VEXAS Syndrome.

OBJECTIVES: VEXAS syndrome is a severe X-linked autoinflammatory disorder caused by somatic mutations in ubiquitin-like modifier activating enzyme 1 (UBA1), with clinical outcomes that vary by UBA1 genotype. We aimed to elucidate genotype-specific inflammatory programs and identify potential therapeutic targets. METHODS: We conducted longitudinal deep phenotyping, including whole-blood RNA sequencing (RNA-seq) and clinical activity assessment. Peripheral blood samples were analyzed by single-cell RNA-seq. Human monocytic cell lines harboring each major UBA1 mutation (p.Met41Val, p.Met41Thr, or p.Met41Leu) were generated and subjected to transcriptomic and functional analyses. RESULTS: Thirteen patients with VEXAS syndrome contributed a total of 79 RNA-seq samples. Among genes upregulated in VEXAS syndrome, RNASE1 showed the strongest correlation with longitudinal disease activity (r = 0.70, FDR < 0.05) and was upregulated in patients' monocytes. In UBA1-mutant monocytic cell lines, genotype-dependent ubiquitination defects were observed in a graded manner (p.Met41Val > p.Met41Thr > p.Met41Leu), even in the absence of exogenous stimuli. These defects were accompanied by unfolded protein response activation, increased pro-inflammatory cytokine production, progressive cell death, and RNASE1 upregulation, all following the same graded pattern, recapitulating patient genotype-phenotype associations. Transcriptomic analyses demonstrated enrichment of pro-inflammatory, interferon, and necroptosis signatures in more severe genotypes. Notably, inhibition of receptor-interacting protein kinase 3 (RIPK3) markedly attenuated all pathological features, including RNASE1 upregulation. CONCLUSIONS: Our UBA1-mutant monocytic cell-line models, representing three distinct genotypes, recapitulate genotype-dependent inflammatory phenotypes that can be modulated by RIPK3 inhibition, providing a translational platform for mechanistic investigation and precision therapy development in VEXAS syndrome.

Journal Article↗

PubMed2005

Modeling the mutualistic interactions between tubeworms and microbial consortia.

The deep-sea vestimentiferan tubeworm Lamellibrachia luymesi forms large aggregations at hydrocarbon seeps in the Gulf of Mexico that may persist for over 250 y. Here, we present the results of a diagenetic model in which tubeworm aggregation persistence is achieved through augmentation of the supply of sulfate to hydrocarbon seep sediments. In the model, L. luymesi releases the sulfate generated by its internal, chemoautotrophic, sulfide-oxidizing symbionts through posterior root-like extensions of its body. The sulfate fuels sulfate reduction, commonly coupled to anaerobic methane oxidation and hydrocarbon degradation by bacterial-archaeal consortia. If sulfate is released by the tubeworms, sulfide generation mainly by hydrocarbon degradation is sufficient to support moderate-sized aggregations of L. luymesi for hundreds of years. The results of this model expand our concept of the potential benefits derived from complex interspecific relationships, in this case involving members of all three domains of life.

Animals↗

PubMed1997

Transgenic mice with increased expression of vascular endothelial growth factor in the retina: a new model of intraretinal and subretinal neovascularization.

Vascular endothelial growth factor (VEGF) has been implicated in retinal neovascularization (NV), but it has been difficult to produce retinal NV with exogenous VEGF. We investigated the effect of increased VEGF expression in the retina using tissue-specific, gain-of-function transgenic mice in which the bovine rhodopsin promoter is coupled to the gene for human VEGF. Three founder mice were obtained and used to generate transgenic lines. One of the lines shows increased expression of VEGF in the retina by reverse transcription coupled to polymerase chain reaction and Northern blots, and the VEGF is localized to photoreceptors by immunohistochemistry. These mice demonstrate new vessels originating from the deep capillary bed of the retina that extend beneath the photoreceptor layer into the subretinal space where they form clumps of blood vessels surrounded by proliferated retinal pigmented epithelial cells. The appearance is similar to subretinal NV seen in some patients, except that the blood vessels originate from the retinal vasculature rather than the choroidal vasculature. One of the other two lines of mice did not show increased expression of VEGF and did not have NV; the other line showed retinal degeneration. This study demonstrates that over-expression of VEGF in the retina is sufficient to cause intraretinal and subretinal NV and provides a valuable new animal model.

Animals↗

PubMed2000

Novel reactions catalysed by antibodies.

New structural data on nonhydrolytic antibody catalysts gained over the past two years confirm that antibodies elicited against transition-state analogues function by differential stabilisation of the transition-state over the ground state through electrostatic, van der Waals, cation-pi and hydrogen-bonding interactions. The lack of chemical catalysis correlates with the low catalytic efficiency. Novel strategies that precisely position a key functional residue in the antibody catalyst combining site have therefore emerged, as demonstrated by crystallographic studies. Whereas antibodies with a bulky residue at position H100c of hypervariable loop H3 adopt different cavity shapes, other antibodies share a common deep combining site. This structural restriction might reflect the use of similar hydrophobic haptens to generate the antibody; novel hapten design or new immunisation strategies may, in the future, lead to more structurally diversified active sites.

Antibodies, Catalytic↗

PubMed2001

Crystal structure of methylmalonyl-coenzyme A epimerase from P. shermanii: a novel enzymatic function on an ancient metal binding scaffold.

BACKGROUND: Methylmalonyl-CoA epimerase (MMCE) is an essential enzyme in the breakdown of odd-numbered fatty acids and of the amino acids valine, isoleucine, and methionine. Present in many bacteria and in animals, it catalyzes the conversion of (2R)-methylmalonyl-CoA to (2S)-methylmalonyl-CoA, the substrate for the B12-dependent enzyme, methylmalonyl-CoA mutase. Defects in this pathway can result in severe acidosis and cause damage to the central nervous system in humans. RESULTS: The crystal structure of MMCE from Propionibacterium shermanii has been determined at 2.0 A resolution. The MMCE monomer is folded into two tandem betaalphabetabetabeta modules that pack edge-to-edge to generate an 8-stranded beta sheet. Two monomers then pack back-to-back to create a tightly associated dimer. In each monomer, the beta sheet curves around to create a deep cleft, in the floor of which His12, Gln65, His91, and Glu141 provide a binding site for a divalent metal ion, as shown by the binding of Co2+. Modeling 2-methylmalonate into the active site identifies two glutamate residues as the likely essential bases for the epimerization reaction. CONCLUSIONS: The betaalphabetabetabeta modules of MMCE correspond with those found in several other proteins, including bleomycin resistance protein, glyoxalase I, and a family of extradiol dioxygenases. Differences in connectivity are consistent with the evolution of these very different proteins from a common precursor by mechanisms of gene duplication and domain swapping. The metal binding residues also align precisely, and striking structural similarities between MMCE and glyoxalase I suggest common mechanisms in their respective epimerization and isomerization reactions.

Binding Sites↗

PubMed2000

Using antibody catalysis to study the outcome of multiple evolutionary trials of a chemical task.

Catalytic aldolase antibodies generated by immunization with two different, but structurally related, beta-diketone haptens were cloned and sequenced to study similarities and differences between independently evolved catalysts. Kinetic and sequence analysis coupled with mutagenesis, structural, and modeling studies reveal that the defining event in the evolution of these catalysts was a somatic mutation that placed a lysine residue in a deep, yet otherwise unrefined, hydrophobic pocket. We suggest that covalent chemistries may be as readily selected from the immune repertoire as the traditional noncovalent interactions that have formed the basis of immunochemistry until this time. Further, we believe that these experiments recapitulate the defining events in the evolution of nature's enzymes, particularly as they relate to chemical mechanism, catalytic promiscuity, and gene duplication.

Amino Acid Sequence↗

PubMedDate not supplied

A current source density analysis of evoked responses in slices of adult rat visual cortex: implications for the regulation of long-term potentiation.

In slices of visual cortex, long-term potentiation (LTP) of synaptic responses in layer III can be evoked by high-frequency stimulation of a site in the middle of the cortical thickness, corresponding mainly to layer IV. In contrast, stimulation of the white matter-layer VI border typically fails to evoke LTP in adult visual cortex unless GABAA receptors are partially blocked. We performed current-source density (CSD) analysis to determine how the patterns of cortical activation compare under these different stimulation conditions. Single-pulse stimulation of the middle layers (corresponding to layer IV and superficial V) and the deep layers (corresponding to white matter and deep layer VI) yielded very similar CSD patterns. The major current sinks were located within 500 mu m of the pia, corresponding to layers II and III, regardless of the stimulation site. The amplitude of all current sinks was diminished, and the latency was increased, in the presence of high concentrations of divalent cations (12 mM Ca2+ and 12 mM mg2+). Nonetheless, the major synaptic current sink was still present at a depth of approximately 400 microns regardless of the site of stimulation, indicating that stimulation of either site leads to monosynaptic EPSCs in layer III. However, superficial sinks, at a depth of approximately 200 microns, were virtually eliminated by high concentrations of divalent cations after deep layer stimulation, but not after middle layer stimulation, suggesting that stimulation at the two sites recruits different monosynaptic circuits. This conclusion was supported by experiments using paired-pulse stimulation of the two sites (12.5 ms interstimulus interval). While there was little evidence of a paired-pulse interaction after stimulation of the middle layers, there was marked paired-pulse suppression of superficial layer III current sinks after stimulation of the deep layers. Taken together, the data suggest a model in which deep layer stimulation activates the dendrites of layer III cells by a monosynaptic route and by a disynaptic route. The disynaptic input originates in the middle cortical layers and is controlled by inhibition. Differences in synaptic plasticity evoked from the different sites could be explained if the recruitment of middle layer inputs were required for the generation of LTP in layer III.

Animals↗

PubMed2026

Using deep learning models as a genetic architecture for the simulation of breeding schemes.

In several simulation studies, long-term selection led to the rapid depletion of genetic variance. These outcomes differ from real-life observations that we aim to replicate, thereby highlighting a fundamental limitation of current classical quantitative genetic simulation models. Deep learning (DL) models have demonstrated promising results in capturing complex interactions essential for maintaining genetic variance; thus, we hypothesize that DL-based genetic simulation models may preserve more genetic variance than classical models, because the biological pathways underlying complex traits exhibit interactions that classical models ignore. The primary objective of this study was to introduce alternative DL-based genetic simulation models and compare them with classical genetic simulation models in terms of their retention of additive genetic variance under truncation selection in a simulated full-sib pig breeding scheme using real haplotypes as founders. After 20 generations of directional truncation selection, the classical models (A, ADAA, and ADAAADDD) retained between 55% and 64% of their initial additive genetic variance. In contrast, while the DL_simple model lost all its additive variance, the DL medium retained 92% to 98% of its additive variance, and the DL_complex model's initial additive variance increased by 296% to 314%. This paper introduces DL-based genetic simulation models and concludes that their ability to retain additive genetic variance depends on the models' architectural complexity. When sufficiently complex, DL-based models exhibit greater retention of additive genetic variance because they intrinsically capture epistatic interactions that are converted into additive variance, as selection progresses, thus, affirming the role of non-additive genetic effects in maintaining long-term genetic variation.

Deep Learning↗

PubMed2005

Granular packings sheared in an annular channel: flow localization and grain size dependence.

We investigate experimentally a quasistatic flow of glass beads in an annular channel, in which particles are packed and sheared from above under a constant normal load. The experiments utilize techniques of refractive-index-matched fluorescent imaging to determine the motion of individual particles and the velocity fields inside the sheared packing. We demonstrated in a previous paper [Phys. Rev. E 70, 031303 (2004)] that an ordering transition has a significant impact on the velocity profile. Here, we report the effects of layer thickness, channel width, and particle size on the internal velocity field. For very thin layers, the grain velocity exhibits a linear vertical profile. As the layer thickness increases, a strongly nonlinear velocity profile emerges, with particle motion that is largely localized to a narrow region (shear band) near the driving surface. Once the packing has reached its steady state, the velocity field is insensitive to the size of grains being used--the velocity profile does not scale with grain size. However, the vertical decay of grain velocity becomes significantly steeper as the horizontal width of the channel decreases. In addition, we demonstrate that changing the direction of shearing generates an anomalous mobility of grains in the deep interior that is sensitive to particle size. The transient grain motion is accompanied by an abrupt volume compaction and a gradual recovery as the shearing proceeds. Reviewing results from this and other works reveals that the velocity profiles of granular shear flows are often geometry specific. We present a heuristic continuum model that qualitatively captures the shear banding observed in this geometry.

Journal Article↗

PubMed2002

The influence of cervical lesion on the stress state and strength of tooth with occlusal restoration: a numerical model study.

This paper is concerned with bio-engineering problems in dentistry. The analysis of stress in a lower premolar with a deep non-carious cervical lesion and restored occlusal Class I cavity under occlusal load of 100 N was shown. Finite element method was used for the analysis of a two-dimensional model of a human tooth. In order to facilitate the evaluation of the loads, four cross-sections in characteristical points of the model were made. The model was generated and the calculations were done on IBM Pentium III personal computer. The principal and equivalent stresses were calculated and used in qualitative and quantative evaluation of tooth tissue and tooth restoration strength. The results are shown in figures and charts. The analysis of the cross-sections distant to the cervical area has proved that the presence or absence of cervical lesion has no significant influence on the values and stress distribution. Only in the region of sharp-shaped notch in the bottom of defect (4-4) high stresses in the tooth tissues occur. The values of stress did not exceed the compressive and tensile strength in any case under investigation. The study has shown that the presence of cervical angular lesion does not affect the mechanical strength of a tooth with a properly restored occlusal cavity. Clinically, the situation is safe from the biomechanical point of view.

Biomechanical Phenomena↗

PubMed1989

[Moderate deficiency of Factor XII associated with postoperative deep venous thrombosis].

The interrelationship between factor XII deficiency (Hageman trait) and thrombosis is well known. A case of moderate factor XII deficiency (activity, 30%) associated to deep vein thrombosis, which occurred in the popliteal region of the left lower limb after abdominal surgery, is reported. The deficit was found in 4 family members of the three generations studied, and all of them showed a close interrelationship between factor XII activity and kallikrein levels. Prolonged APTT was found in 3 of the 4 affected subjects. A multiallelic model is suggested to explain the genetic transmission of this impairment.

Aged↗

PubMed2000

Experimental models of growth factor-mediated angiogenesis and blood-retinal barrier breakdown.

Following chronic ischemia, vascular endothelial growth factor (VEGF) is induced primarily in the ganglion cell layer of the retina. This often results in neovascularization (NV) that originates from the vascular bed closest to the ganglion cell layer. To study the effects of VEGF, independent lines of transgenic mice that express VEGF in the lens and in the retina have been generated. Expression in the lens results in excessive proliferation and accumulation of angioblasts and endothelial cells in proximity to the lens. However, VEGF expression is not sufficient to direct blood vessel organization or maturation in the prenatal mouse. Abnormal vessels do form on the retinal surface, but not until the second postnatal week. In transgenic mice expressing VEGF in the photoreceptors, NV originates from the deep capillary bed--the vascular bed closest to the photoreceptors. NV is accompanied by localized blood-retinal barrier breakdown. NV is also induced in PDGF-B transgenic mice. PDGF-B expression in the lens occurs prenatally and, during this time, mainly affects the perilenticular vessels. Postnatally, transgenic mice expressing PDGF-B in the lens or photoreceptors show a similar phenotype. In both models, a highly vascularized cell mass containing endothelial cells, pericytes, and glia forms in the superficial retina, and the formation of the deep capillary bed is inhibited. The phenotype suggests that an additional factor is necessary for the maturation and penetration of vascular endothelial cells into the retina to form the deep capillary bed.

Animals↗

PubMed2006

Skin wound healing in the first generation (F1) offspring of Yorkshire and red Duroc pigs: evidence for genetic inheritance of wound phenotype.

Fibroproliferative scars in humans often demonstrate familial inheritance patterns, and genetics may contribute to healing and scarring. Genetic factors may also influence the scarring phenotype in a porcine model. Healing of full thickness excisional skin wounds in Yorkshire pigs closely resembles normal healing in humans, while identical wounds in red Duroc pigs form hypercontracted, hyperpigmented scars. The present study has evaluated the healing process in the first generation cross (F1) of red Duroc and Yorkshire pigs. Gross and histologic analysis revealed that the F1 animals exhibit an intermediate healing phenotype, with some features of each parent breed. F1 full thickness wounds were significantly hypercontracted and fibrotic, but apigmented. Analysis of mRNA expression patterns for a panel of relevant molecules (N=32) in the F1 animals revealed some similarities to each parent breed, as well as unique patterns for other molecules. Furthermore, a depth dependency to the healing response was observed at the gross, histologic, and molecular levels, with deep dermal wounds healing similar to Yorkshire wounds. These findings suggest that the genetic contribution to scar phenotype in this animal model is complex. However, the results indicate that further understanding in this model may provide insights into risk factors for hypertrophic scarring in human burn patients.

Animals↗

PubMed2000

Establishment of an animal-bacterial association: recruiting symbiotic vibrios from the environment.

While most animal-bacterial symbioses are reestablished each successive generation, the mechanisms by which the host and its potential microbial partners ensure tissue colonization remain largely undescribed. We used the model association between the squid Euprymna scolopes and Vibrio fischeri to examine this process. This light organ symbiosis is initiated when V. fischeri cells present in the surrounding seawater enter pores on the surface of the nascent organ and colonize deep epithelia-lined crypts. We discovered that when newly hatched squid were experimentally exposed to natural seawater, the animals responded by secreting a viscous material from the pores of the organ. Animals maintained in filtered seawater produced no secretions unless Gram-negative bacteria, either living or dead, were reintroduced. The viscous material bound only lectins that are specific for either N-acetylneuraminic acid or N-acetylgalactosamine, suggesting that it was composed of a mucus-containing matrix. Complex ciliated fields on the surface of the organ produced water currents that focused the matrix into a mass that was tethered to, and suspended above, the light organ pores. When V. fischeri cells were introduced into the seawater surrounding the squid, the bacteria were drawn into its fluid-filled body cavity during ventilation and were captured in the matrix. After residing as an aggregate for several hours, the symbionts migrated into the pores and colonized the crypt epithelia. This mode of infection may be an example of a widespread strategy by which aquatic hosts increase the likelihood of successful colonization by rarely encountered symbionts.

Animals↗

PubMed2023

Towards mechanistic models of mutational effects: Deep learning on Alzheimer's Aβ peptide.

Deep Mutational Scanning (DMS) has enabled multiplexed measurement of mutational effects on protein properties, including kinematics and self-organization, with unprecedented resolution. However, potential bottlenecks of DMS characterization include experimental design, data quality, and depth of mutational coverage. Here, we apply deep learning to comprehensively model the mutational effect of the Alzheimer's Disease associated peptide Aβ42 on aggregation-related biochemical traits from DMS measurements. Among tested neural network architectures, Convolutional Neural Networks and Recurrent Neural Networks are found to be the most cost-effective models with high performance even under insufficiently-sampled DMS studies. While sequence features are essential for satisfactory prediction from neural networks, geometric-structural features further enhance the prediction performance. Notably, we demonstrate how mechanistic insights into phenotype may be extracted from the neural networks themselves suitably designed. This methodological benefit is particularly relevant for biochemical systems displaying a strong coupling between structure and phenotype such as the conformation of Aβ42 aggregate and nucleation, as shown here using a Graph Convolutional Neural Network (GCN) developed from the protein atomic structure input. In addition to accurate imputation of missing values (which here ranged up to 55% of all phenotype values at key residues), the mutationally-defined nucleation phenotype generated from a GCN shows improved resolution for identifying known disease-causing mutations relative to the original DMS phenotype. Our study suggests that neural network derived sequence-phenotype mapping can be exploited not only to provide direct support for protein engineering or genome editing but also to facilitate therapeutic design with the gained perspectives from biological modeling.

Alzheimer's disease↗

PubMed1996

Molecular modeling of protocellular functions.

The mechanisms of three protocellular functions have been studied using molecular modeling techniques. These functions are (1) the transport of ions across membranes, (2) the formation of photoactivated proton gradient that could drive chemical synthesis in the protocell, and (3) the organization of small peptides necessary for catalytic activity. In all these processes, membranes play an essential role. The transfer of ions across the barrier formed by protocellular walls is facilitated by the formation of deep, thinning defects in the membrane. Membranes also form a barrier to charged species that allows for retaining proton gradients. These gradients can be generated by a simple transmembrane proton pump consisting of a proton source and two acceptors. The directionality of the pump is ensured by a "gate-keeping" mechanism involving a water molecule, conformational change of the primary acceptor or tautomerization of a histidine. The pump can be formed by two transmembrane helices but not one helix. They provide surfaces at which organic molecules concentrate and small peptides can organize into ordered, amphiphilic structures. In general, valuable information about the origins and evolution of protocells can be obtained from the knowledge of physical and chemical principles that govern functioning of contemporary cells.

Amino Acid Sequence↗