SEARCH · PubMed Health

Results for “Predictive Learning 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 1,135 records · Page 63Linked to original sources

PubMed2025

DeepWheat: predicting the effects of genomic variants on gene expression and regulatory activities across tissues and varieties in wheat using deep learning.

Spatiotemporal gene expression shapes key agronomic traits, yet tissue-specific prediction remains challenging in complex crops. We present DeepWheat, a broadly applicable deep learning framework comprising DeepEXP and DeepEPI, for accurate, tissue-specific gene expression prediction. DeepEXP integrates sequence and epigenomic features to predict gene expression (PCC 0.82-0.88), while DeepEPI predicts epigenomic maps from DNA sequence to support model transfer across varieties. Validations in five wheat cultivars confirm robustness and accuracy. DeepWheat also identifies regulatory variants with strong expression effects, enabling targeted cis-regulatory elements editing and offering a powerful tool for crop functional genomics and breeding.

Triticum↗

PubMed2026

Integrating structure and experimental data annotations with computational modeling framework for predicting micro-nanoplastics toxicities.

The wide use of plastic materials leads to increased emissions of micro-nanoplastics (MNPs) into the environment, raising significant concerns about their impact on human health. Traditional experimental approaches for assessing MNPs toxicity are costly, time-consuming, and there are no experimental protocols that are universally acceptable. Computational modeling using machine learning (ML) approaches provides an efficient alternative to MNP toxicity assessment. However, most modeling studies of MNPs are limited due to the lack of high-quality data and there are few previous modeling studies considering complex structures of MNPs for model training. To address this challenge, we constructed three MNP datasets with popular toxicity endpoints from various resources and used nanostructure annotation techniques to create virtual MNPs (vMNPs) for all MNP structures. The MNP structures were digitalized from annotated vMNPs, and geometrical descriptors were calculated using the Delaunay Tessellation approach. Moreover, important experimental information, such as concentrations and cell lines, were transformed into extra training variables. Partial least squares regression (PLSR) models were built using both experimental and geometrical descriptors and validated through a leave-one-out cross validation procedure. The resulting models showed reasonable performance in predicting toxicity potentials of MNPs for the three endpoints in the present datasets. Moreover, an additional library of vMNPs with their predicted properties and bioactivities was constructed, directing further research of new MNPs. This study provides three novel ML models for MNPs by integrating geometrical and experimental descriptors, which have the potential to assess new MNPs for their toxicity. The modeling strategy developed in this study can be easily expanded to model other MNP toxicity endpoints and create promising new models for MNP toxicity assessments.

Data annotation↗

PubMedDate not supplied

Aging and classical conditioning: parallel studies in rabbits and humans.

The model system of classical conditioning of the eyelid response in rabbits has led to significant progress in the understanding of the neural circuitry involved in learning and memory. We are now at a point where we can begin to explore how aging in the neural circuitry affects this basic form of associative learning and memory over the life span. Parallels between humans and animals in aging and classical conditioning of the eyelid response are identified. A hypothetical model describing how acquisition occurs in the cerebellum has proved useful in predicting where in the cerebellum age changes might be most likely to affect classical conditioning. The loss of Purkinje cells with age is one of the cerebellar age changes likely to affect classical conditioning in aged mammals. Purkinje cell loss is only one of several age changes in the cerebellum which may impair acquisition and retention of the conditioned eyeblink response. This model system is of demonstrated utility in extending understanding of the neurobiology of learning, memory, and aging in humans as well as animals.

Adult↗

PubMed1994

Development of acute tolerance after oral doses of diazepam and flunitrazepam.

Flunitrazepam (1 and 2 mg), diazepam (10 and 20 mg) or placebo was administered to healthy, male volunteers, and the time course of psychomotor impairment, as indicated by simple and complex choice reaction time and movement time, was studied during a period of 6 h after drug intake. To examine whether acute tolerance developed, the observed performance during decreasing drug plasma concentration was compared to the predicted performance based on kinetic-dynamic modelling of the observed performance during the first 1.5 h after intake when the drug plasma concentrations were increasing or at peak level. Placebo corrections of the test scores were accomplished to adjust for diurnal variation and the possible influence of learning during the test day. After the flunitrazepam treatments, the predictions overestimated the actual performance significantly with respect to simple and choice reaction time at the 6-h session after intake. After the diazepam treatments, however, no significant deviation was detected between predicted and observed performance. The results indicate that acute tolerance develops with respect to impairment of attention demanding performance after medium to large doses of flunitrazepam, and that tolerance is expressed after approximately 4-6 h following intake.

Adult↗

PubMed2026

Spectral-Proteomic Integration Analysis (SPIA) Deciphers Molecular Trajectories of Breast Cancer and Enables Multitarget Therapeutic Assessment.

Raman spectroscopy and mass spectrometry-based proteomics offer deeply complementary yet largely disconnected views of cancer biology: the former provides a label-free, real-time biochemical phenotype, while the latter delivers a quantitative inventory of specific protein effectors. Bridging this gap remains a fundamental challenge in analytical biomedicine. Here, we introduce Spectral-Proteomic Integration Analysis (SPIA)─a novel, data-driven integrative framework that systematically links Raman spectroscopic phenotypes with quantitative proteomic profiles through machine learning and statistical correlation. Using a DMBA-induced rat breast cancer model with and without Toremifene (TOR) intervention, SPIA dynamically maps tumor microenvironment remodeling, capturing progressive collagen deposition and lipid metabolic reprogramming. An SVM classifier trained on Raman spectra achieves exceptional diagnostic accuracy (AUC ≥ 99.0%) and successfully predicts TOR therapeutic response. Proteomic analysis identifies 1,350 differentially expressed proteins, with convergent machine learning feature selection (LASSO, Random Forest, XGBoost) pinpointing core regulators including Luc7l2, Nucb1, Cbx3, and Csnk2a1. Crucially, Spearman correlation analysis between key Raman bands and core DEPs reveals strong, statistically robust associations (median ρ ∼ 0.75 in the 1533-1669 cm-1 region), empirically validating SPIA's core integrative logic. Leveraging this multimodal map, we elucidate a multitarget mechanism for TOR involving concurrent suppression of collagen deposition and correction of aberrant lipid metabolism. SPIA establishes a powerful, generalizable paradigm for integrating phenotypic and molecular data, with broad implications for biomarker discovery, drug mechanism elucidation, and precision oncology.

Animals↗

PubMed2005

Predictors of antral follicle count during the reproductive years.

BACKGROUND: We sought to identify indicators of antral follicle count which would be serviceable to clinicians seeking to estimate the number of ovarian follicles without relying on sonographic counts. METHODS: We examined the relations of chronological age and four potential indicators of ovarian age-ovarian volume, FSH, dimeric inhibin B and estradiol-to antral follicle count in 176 recently pregnant women. We identified the regression models which best predict low antral follicle count (< or =10 follicles). RESULTS: Chronological age, ovarian volume, FSH and inhibin B were each significantly associated with antral follicle count. Fifty-three (30.1%) women had < or =10 antral follicles. In the total sample, at the cutpoint corresponding to 80% sensitivity, the positive predictive value for a regression model with all four variables was 60%. All regression models performed less well in women <35 years (13.9% with low count) than in women > or =35 years (52.0% with low count). In older women, the positive predictive value for the model with all four variables was 79%, compared with 60% for a model with chronological age alone. CONCLUSIONS: Our models provide a basis for advising women aged > or =35 years who are either trying to conceive or wish to learn whether they may postpone childbearing.

Adult↗

PubMed2006

EEG markers for cognitive decline in elderly subjects with subjective memory complaints.

New treatments for Alzheimer's disease require early detection of cognitive decline. Most studies seeking to identify markers of early cognitive decline have focused on a limited number of measures. We sought to establish the profile of brain function measures which best define early neuropsychological decline. We compared subjects with subjective memory complaints to normative controls on a wide range of EEG derived measures, including a new measure of event-related spatio-temporal waves and biophysical modeling, which derives anatomical and physiological parameters based on subject's EEG measurements. Measures that distinguished the groups were then related to cognitive performance on a variety of learning and executive function tasks. The EEG measures include standard power measures, peak alpha frequency, EEG desynchronization to eyes-opening, and global phase synchrony. The most prominent differences in subjective memory complaint subjects were elevated alpha power and an increased number of spatio-temporal wave events. Higher alpha power and changes in wave activity related most strongly to a decline in verbal memory performance in subjects with subjective memory complaints, and also declines in maze performance and working memory reaction time. Interestingly, higher alpha power and wave activity were correlated with improved performance in reverse digit span in the subjective memory complaint group. The modeling results suggest that differences in the subjective memory complaint subjects were due to a decrease in cortical and thalamic inhibitory gains and slowed dendritic time-constants. The complementary profile that emerges from the variety of measures and analyses points to a nonlinear progression in electrophysiological changes from early neuropsychological decline to late-stage dementia, and electrophysiological changes in subjective memory complaint that vary in their relationships to a range of memory-related tasks.

Aged↗

PubMed1994

ECG myocardial infarct size: a gender-, age-, race-insensitive 12-segment multiple regression model. I: Retrospective learning set of 100 pathoanatomic infarcts and 229 normal control subjects.

In this early study of ongoing work with multiple regression modeling for mapping myocardial infarct (MI) into 12 left ventricular (LV) segments, promising results have been presented using electrocardiographic (ECG) QRS variables that are gender, age, and race insensitive (GARI), the GARI-QRS 12-segment multiple regression model. These include Q, R, and S duration, expressed as percentage total QRS duration, and R/Q duration, R/Q amplitude, R/S duration, and R/S amplitude variables. For version I, building 12 regression models using 68 single and 32 multiple MIs, the GARI-QRS variables correlated with pathoanatomic MI in each of 12 segments with r values ranging from .67 to .88. In version II of the model, using all MIs and 229 normal subjects, r = .73-.91. Version II predictions of MI in 12 LV segments for each subject were used to calculate the predicted total percentage LV infarct, which correlated well with that found at autopsy. The r values found were .81 for all single MIs, .73 for multiple MIs, and .80 for all MIs taken together. With refinements of the input ECG variables to include (1) improvement in the GARI-QRS variables, (2) adding a significant number of subjects with hypertrophies and conduction defects with and without MI to an expanded learning set, and (3) applying the enhanced 12-LV-segment regression models to a similar test set, it is to be expected that these regression models can be improved even further in such a way as to be applicable to general clinical populations using routine computerized ECG analysis programs.

Adult↗

PubMed2006

Learning by the parasitoid wasp, Aphidius ervi (Hymenoptera: Braconidae), alters individual fixed preferences for pea aphid color morphs.

Learning, defined as changes in behavior that occur due to past experience, has been well documented for nearly 20 species of hymenopterous parasitoids. Few studies, however, have explored the influence of learning on population-level patterns of host use by parasitoids in field populations. Our study explores learning in the parasitoid Aphidius ervi Haliday that attacks pea aphids, Acyrthosiphon pisum (Harris). We used a sequence of laboratory experiments to investigate whether there is a learned component in the selection of red or green aphid color morphs. We then used the results of these experiments to parameterize a model that examines whether learned behaviors can explain the changes in the rates of parasitism observed in field populations in South-central Wisconsin, USA. In the first of two experiments, we measured the sequence of host choice by A. ervi on pea aphid color morphs and analyzed this sequence for patterns in biased host selection. Parasitoids exhibited an inherent preference for green aphid morphs, but this preference was malleable; initial encounters with red aphids led to a greater chance of subsequent orientation towards red aphids than predicted by chance. In a second experiment, we found no evidence that parasitoids specialize on red or green morphs; for the same parasitoids tested in trials separated by 2 h, color preference in the first trial did not predict color preference in the second, as would be expected if they differed in fixed preferences or exhibited long-term (> 2 h) learning. Using data from the two experiments, we parameterized a population dynamics model and found that learning of the magnitude observed in our experiments leads to biased parasitism towards the most common color morph. This bias is sufficient to explain changes in the ratio of aphid color morphs observed in field sites over multiple years. Our study suggests that for even relatively simple organisms, learned behaviors may be important for explaining the population dynamics of their hosts.

Animals↗

PubMed1982

Operant control of turning in circles: a new model of dopaminergic drug action.

Rats were trained, using water reinforcement, to exhibit high rates of rotation (circling) during one-hour daily test sessions. Preferred directions of learned rotation were the same as those determined previously in response to D-amphetamine. Changes in reinforcement parameters elicited predictable changes in rates of learned rotation. The effects of D-amphetamine, apomorphine, haloperidol and methohexital could be readily dissociated indicating that the operant rotation paradigm could likely become a useful behavioral assay of dopaminergic drug action.

Animals↗

PubMed2001

Learned helplessness in the rat: improvements in validity and reliability.

Major depression has a high prevalence and a high mortality. Despite many years of research little is known about the pathophysiologic events leading to depression nor about the causative molecular mechanisms of antidepressant treatment leading to remission and prevention of relapse. Animal models of depression are urgently needed to investigate new hypotheses. The learned helplessness paradigm initially described by Overmier and Seligman [J. Comp. Physiol. Psychol. 63 (1967) 28] is the most widely studied animal model of depression. Animals are exposed to inescapable shock and subsequently tested for a deficit in acquiring an avoidance task. Despite its excellent validity concerning the construct of etiology, symptomatology and prediction of treatment response [Clin. Neurosci. 1 (1993) 152; Trends Pharmacol. Sci. 12 (1991) 131] there has been little use of the model for the investigation of recent theories on the pathogenesis of depression. This may be due to reported difficulties in reliability of the paradigm [Animal Learn. Behav. 4 (1976) 401; Pharmacol. Biochem. Behav. 36 (1990) 739]. The aim of the current study was therefore to improve parameters for inescapable shock and learned helplessness testing to minimize artifacts and random error and yield a reliable fraction of helpless animals after shock exposure. The protocol uses mild current which induces helplessness only in some of the animals thereby modeling the hypothesis of variable predisposition for depression in different subjects [Psychopharmacol. Bull. 21 (1985) 443; Neurosci. Res. 38 (200) 193]. This allows us to use animals which are not helpless after inescapable shock as a stressed control, but sensitivity, specificity and variability of test results have to be reassessed.

Animals↗

PubMed2000

Modeling place fields in terms of the cortical inputs to the hippocampus.

A model of place-cell firing is presented that makes quantitative predictions about specific place cells' spatial receptive fields following changes to the rat's environment. A place cell's firing rate is modeled as a function of the rat's location by the thresholded sum of the firing rates of a number of putative cortical inputs. These inputs are tuned to respond whenever an environmental boundary is at a particular distance and allocentric direction from the rat. The initial behavior of a place cell in any environment is simply determined by its set of inputs and its threshold; learning is not necessary. The model is shown to produce a good fit to the firing of individual place cells, and populations of place cells across environments of differing shape. The cells' behavior can be predicted for novel environments of arbitrary size and shape, or for manipulations such as introducing a barrier. The model can be extended to make behavioral predictions regarding spatial memory.

Afferent Pathways↗

PubMed2006

fMRI models of dendritic and astrocytic networks.

In order to elucidate the relationships between hierarchical structures within the neocortical neuropil and the information carried by an ensemble of neurons encompassing a single voxel, it is essential to predict through volume conductor modeling LFPs representing average extracellular potentials, which are expressed in terms of interstitial potentials of individual cells in networks of gap-junctionally connected astrocytes and synaptically connected neurons. These relationships have been provided and can then be used to investigate how the underlying neuronal population activity can be inferred from the measurement of the BOLD signal through electrovascular coupling mechanisms across the blood-brain barrier. The importance of both synaptic and extrasynaptic transmission as the basis of electrophysiological indices triggering vascular responses between dendritic and astrocytic networks, and sequential configurations of firing patterns in composite neural networks is emphasized. The purpose of this review is to show how fMRI data may be used to draw conclusions about the information transmitted by individual neurons in populations generating the BOLD signal.

Animals↗

PubMed1997

Connectionism, phonology, reading, and regularity in developmental dyslexia.

Tests of the "phonological deficit" account of developmental dyslexia have produced apparently inconsistent results. We show how a connectionist approach to dyslexic reading development can resolve the paradox. A "dyslexic" model of reading was created by reducing the quality of the phonological representations available to the model during learning. The model behaved similarly to dyslexic children in that it had a selectively reduced ability to process nonwords, but showed normal effects of words' spelling-to-sound regularity. An experimental test of the model's predictions confirmed that dyslexic children perform similarly, in that they are impaired on irregular words to the same extent as nondyslexic children. It is concluded that developmentally dyslexic reading can indeed be understood in terms of impaired phonological representations and that the adoption of a modeling approach resolves an apparent paradox in the experimental literature.

Child↗

PubMed1999

Rapid analysis of the expression of heterologous proteins in Escherichia coli using pyrolysis mass spectrometry and Fourier transform infrared spectroscopy with chemometrics: application to alpha 2-interferon production.

Cell pastes and supernatant Escherichia coli samples, taken from an industrial bioprocess overproducing recombinant alpha 2 IFN were analysed using pyrolysis mass spectrometry (PyMS) and diffuse reflectance-absorbance Fourier transform infrared spectroscopy (FT-IR). PyMS and FT-IR are physico-chemical methods which measure predominantly the bond strengths of molecules and the vibrations of bonds within functional groups, respectively. They therefore give quantitative information about the total biochemical composition of the bioprocess sample. The interpretation of these hyperspectral data, in terms of the quantity of alpha 2 IFN in the cell pastes and supernatant samples was possible only after the application of the 'supervised learning' methods of artificial neural networks (ANNs) and partial least squares (PLS) regression. Both PyMS and FT-IR are novel, rapid and economical methods for the screening and the quantitative analysis of complex biological bioprocess over producing recombinant proteins. Models established using either spectral data set had a similarly satisfactory predictive ability. This shows that whole-reaction mixture spectral methods, which measure all molecules simultaneously, do contain enough information to allow their quantification when the entire spectra are used as the inputs to methods based on supervised learning. Moreover, this is the first study where FT-IR in the mid-IR range has been used to quantify the expression of a heterologous protein directly from fermentation broths and the first study to compare the abilities of PyMS and FT-IR for the quantitative analyses of an industrial bioprocess.

Chemistry Techniques, Analytical↗

PubMed2000

Basic concepts of artificial neural network (ANN) modeling and its application in pharmaceutical research.

Artificial neural networks (ANNs) are biologically inspired computer programs designed to simulate the way in which the human brain processes information. ANNs gather their knowledge by detecting the patterns and relationships in data and learn (or are trained) through experience, not from programming. An ANN is formed from hundreds of single units, artificial neurons or processing elements (PE), connected with coefficients (weights), which constitute the neural structure and are organised in layers. The power of neural computations comes from connecting neurons in a network. Each PE has weighted inputs, transfer function and one output. The behavior of a neural network is determined by the transfer functions of its neurons, by the learning rule, and by the architecture itself. The weights are the adjustable parameters and, in that sense, a neural network is a parameterized system. The weighed sum of the inputs constitutes the activation of the neuron. The activation signal is passed through transfer function to produce a single output of the neuron. Transfer function introduces non-linearity to the network. During training, the inter-unit connections are optimized until the error in predictions is minimized and the network reaches the specified level of accuracy. Once the network is trained and tested it can be given new input information to predict the output. Many types of neural networks have been designed already and new ones are invented every week but all can be described by the transfer functions of their neurons, by the learning rule, and by the connection formula. ANN represents a promising modeling technique, especially for data sets having non-linear relationships which are frequently encountered in pharmaceutical processes. In terms of model specification, artificial neural networks require no knowledge of the data source but, since they often contain many weights that must be estimated, they require large training sets. In addition, ANNs can combine and incorporate both literature-based and experimental data to solve problems. The various applications of ANNs can be summarised into classification or pattern recognition, prediction and modeling. Supervised 'associating networks can be applied in pharmaceutical fields as an alternative to conventional response surface methodology. Unsupervised feature-extracting networks represent an alternative to principal component analysis. Non-adaptive unsupervised networks are able to reconstruct their patterns when presented with noisy samples and can be used for image recognition. The potential applications of ANN methodology in the pharmaceutical sciences range from interpretation of analytical data, drug and dosage form design through biopharmacy to clinical pharmacy.

Algorithms↗

PubMed1995

Saccadic motor planning by integrating visual information and pre-information on neural dynamic fields.

A functional model of target selection in the saccadic system is presented, incorporating elements of visual processing, motor planning, and motor control. We address the integration of visual information with pre-information, which is provided by manipulating the probability that a target appears at a certain location. This integration is achieved within a dynamic representation of planned eye movement which is modeled through distributions of activation on a topographic field. Visual input evokes activation, which is also constrained by lateral interaction within the field and by preshaping input representing pre-information. The model describes target selection observable in paradigms in which visual goals are presented at more than one location. Specifically, we model the transition from averaging, where endpoints of first saccades fall between two visual target locations, to decision making, where endpoints of first saccades fall accurately onto one of two simultaneously presented visual targets. We make predictions about how metrical biases of first saccades are induced by pre-information about target locations acquired by learning. When coupled to a motor control stage, activation dynamics on the planning level contribute to stabilizing gaze under fixation conditions. The neurophysiological relevance of our functional model is discussed.

Computer Simulation↗

PubMed2026

Deep learning and statistical methods identify novel asthma risk variants in Europeans.

BACKGROUND: Asthma is a common heritable respiratory disorder with a complex genetic basis. Although large-scale genome-wide association studies have identified many risk loci, the full spectrum of its polygenic architecture remains to be defined. OBJECTIVE: We refined the genetic landscape of asthma in individuals of European ancestry and improve polygenic risk prediction through statistical and deep learning-based methods. METHODS: We conducted the largest genome-wide association study meta-analysis of asthma in individuals of European ancestry, combining data from the Global Biobank Meta-analysis Initiative (121,940 cases, 1,254,131 controls) and the Million Veteran Program (36,823 cases, 398,278 controls). To enhance discovery, we applied pleiotropy-informed multitrait analysis and conditional false discovery rate approaches, each incorporating eosinophil counts as a secondary trait. In parallel, we used a Transformer-based deep learning framework to further prioritize variants and improve polygenic risk prediction. RESULTS: The meta-analysis identified 69 independent genome-wide significant loci (P < 5 × 10-8) not previously reported in asthma. Multitrait analysis of genome-wide association studies, conditional false discovery rate, and deep learning approaches uncovered additional candidate loci. Functional annotation and expression quantitative trait locus mapping implicated novel genes in immune regulation, airway remodeling, and metabolic processes. Polygenic risk score models derived from deep learning-prioritized variants outperformed those based on conventional genome-wide association study and standard statistical approaches. CONCLUSIONS: Our study yields a comprehensive map of asthma-associated loci in European ancestry populations, improves genetic risk prediction, and informs future mechanistic studies.

Humans↗