SEARCH · Search PubMed

Results for “Benchmark”

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.

SourceResearch connectionFrom yearThrough year

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

At least 1,585 records · Page 88Linked to original sources

PubMed2004

"Best practice" in inflammatory bowel disease: an international survey and audit.

Background: An observational study was conducted at eight university and four district hospitals in eight countries collaborating in clinical and epidemiological research in inflammatory bowel disease (IBD) to compare European health care facilities and to define current "best practice" with regard to IBD. Methods: The approach used in this multi-national survey was unique. Existing quality norms, developed for total hospital care by a specialized organization, were restricted to IBD-specific care and adapted to the frame of reference of the study group. In each center, these norms were surveyed by means of questionnaires and professional audits in all participating centers. The collected data were reported to the center, compared to data from other hospitals, and used to benchmark. Group consensus was reached with regard to defining current "best practice". Results: The observations in each center involved patient-oriented processes, technical and patient safety, and quality of the medical standard. Several findings could be directly implemented to improve IBD care in another hospital (benchmarks). These included a confidential relationship between health care worker(s) and patients, and availability of patient data. Conclusions: The observed benchmarks, in combination with other subjectively chosen "positive" procedures, have been defined as current "best practice in IBD", representing practical guidelines towards better quality of care in IBD.

Journal Article↗

PubMed2026

Current Diagnostic Pathways for Rheumatoid Arthritis-Associated Interstitial Lung Disease Result in Substantial Underdiagnosis and Excess Mortality: A Multicenter Norwegian Quality Assurance Audit.

OBJECTIVE: Recent guidelines suggest risk-stratified screening for rheumatoid arthritis-associated interstitial lung disease (RA-ILD). However, the diagnostic gap between current routine care and this screening approach remains unquantified. We assessed currently detected RA-ILD in Norway, benchmarking findings against recent screening-based estimates of the true disease burden. METHODS: This 10-year quality assurance audit across six centers covered 43% of the Norwegian population. RA-ILD cases identified via ICD-10 codes were confirmed by manual chart review. Prevalence was calculated relative to a registry-derived total RA background population and benchmarked against a 10% expected target derived from recent prospective studies. Mortality was compared to a 3:1 frequency-matched RA control group using Cox proportional hazards regression. RESULTS: Among 17,305 RA patients, 188 (1.1%) had verified ILD; when benchmarked against an expected 10% prevalence, this indicates an 89% diagnostic gap in routine clinical care. Mean age at ILD detection was 67.5 years. Most cases (93.6%) possessed ≥2 established risk factors for RA-ILD: 93.6% were seropositive, 76.1% had smoking histories, while RA onset age ≥60 and persistently increased inflammatory laboratory markers were present in over half of patients. RA-ILD was associated with significantly increased mortality; 66 (4.1/100 person-years) deaths occurred in the RA-ILD group vs. 120 (2.3/100 person-years) among RA controls (HR 1.77; 95% CI: 1.31-2.39, p<0.001). CONCLUSION: When comparing to prevalence expectations, current routine care may leave a substantial proportion of cases undetected, primarily capturing a high-risk phenotype with excess mortality. Systematic, risk-stratified screening is needed to bridge this diagnostic gap, aiming to enable earlier intervention.

Interstitial lung disease↗

PubMed2024

metaExpertPro: A Computational Workflow for Metaproteomics Spectral Library Construction and Data-Independent Acquisition Mass Spectrometry Data Analysis.

Analysis of large-scale data-independent acquisition mass spectrometry metaproteomics data remains a computational challenge. Here, we present a computational pipeline called metaExpertPro for metaproteomics data analysis. This pipeline encompasses spectral library generation using data-dependent acquisition MS, protein identification and quantification using data-independent acquisition mass spectrometry, functional and taxonomic annotation, as well as quantitative matrix generation for both microbiota and hosts. By integrating FragPipe and DIA-NN, metaExpertPro offers compatibility with both Orbitrap and timsTOF MS instruments. To evaluate the depth and accuracy of identification and quantification, we conducted extensive assessments using human fecal samples and benchmark tests. Performance tests conducted on human fecal samples indicated that metaExpertPro quantified an average of 45,000 peptides in a 60-min diaPASEF injection. Notably, metaExpertPro outperformed three existing software tools by characterizing a higher number of peptides and proteins. Importantly, metaExpertPro maintained a low factual false discovery rate of approximately 5% for protein groups across four benchmark tests. Applying a filter of five peptides per genus, metaExpertPro achieved relatively high accuracy (F-score = 0.67-0.90) in genus diversity and showed a high correlation (rSpearman = 0.73-0.82) between the measured and true genus relative abundance in benchmark tests. Additionally, the quantitative results at the protein, taxonomy, and function levels exhibited high reproducibility and consistency across the commonly adopted public human gut microbial protein databases IGC and UHGP. In a metaproteomic analysis of dyslipidemia patients, metaExpertPro revealed characteristic alterations in microbial functions and potential interactions between the microbiota and the host.

Proteomics↗

PubMed2002

Reliability of transmembrane predictions in whole-genome data.

Transmembrane prediction methods are generally benchmarked on a set of proteins with experimentally verified topology. We have investigated if the accuracy measured on such datasets can be expected in an unbiased genomic analysis, or if there is a bias towards 'easily predictable' proteins in the benchmark datasets. As a measurement of accuracy, the concordance of the results from five different prediction methods was used (TMHMM, PHD, HMMTOP, MEMSAT, and TOPPRED). The benchmark dataset showed significantly higher levels (up to five times) of agreement between different methods than in 10 tested genomes. We have also analyzed which programs are most prone to make mispredictions by measuring the frequency of one-out-of-five disagreeing predictions.

Algorithms↗

PubMed1997

Nosocomial infections in pediatric patients with burns.

BACKGROUND: Nosocomial infections (NI) are believed to occur more commonly in patients with burns than in patients undergoing surgery, but benchmark rates have not been well described, and widely accepted definitions of NI in patients with burns are not available. We present a clinically useful set of definitions for NI for the pediatric burn population and provide benchmark infection rates for NI at selected sites. METHODS: Centers for Disease Control and Prevention definitions were modified to more accurately describe nosocomial burn infection and secondary bloodstream infections (BSI) in the burn population. A surveillance system was developed and included calculation of NI rates by 1000 patient or device days, stratified into one of three risk groups (< 30% burn injury, 30% to 60% burn injury, and > 60% burn injury). All patients with acute burns admitted from January 1990 to December 1991 were included, and NI rates were calculated for burn infection, primary and secondary BSI, ventilator-related pneumonia and urinary catheter-related urinary tract infection (UTI). RESULTS: Overall 12.5% of patients with central venous catheters had development of primary BSI for a rate of 4.9/1000 central venous catheter-days. Incidence of secondary BSI was 5.8% of patients for a rate of 5.3/1000 patient-days. Incidence of burn infection was 10.1% of patients for a rate of 5.6/1000 patient-days. Incidence of ventilator-related pneumonia was 17.5% of patients for a rate of 11.4/1000 ventilator-days. Incidence of urinary catheter-related UTI was 17.9% of patients, for a rate of 13.2/1000 urinary catheter-days. When rates were stratified by risk groups, incidence increased with increasing burn size for secondary BSI (p < or = 0.0001) and urinary catheter-related UTI (p = 0.08), although rates based on number of patient-days or device-days more accurately reflected risk of infection over time. CONCLUSIONS: Infection remains a cause of significant morbidity and death for patients with burns. The definitions and benchmark rates reported here may be useful in evaluation of NI surveillance strategies and calculation of infection rates, which could then be used to evaluate current treatment modalities and improve outcomes for the burn population.

Burns↗

PubMed2003

Utilizing data from multiple studies (meta-analysis) to determine effective dose-duration levels. Example: rats and mice exposed to hydrogen sulfide.

The objective of this exercise was to incorporate as much data as possible from multiple studies, that may differ in exposure durations, to derive a chemical-specific dose-duration response curve from which to identify toxicity markers (e.g., ED01, benchmark dose, and LD50). This has the advantage of incorporating more information than single-study assessments to improve estimates and reduce confidence intervals, and determining toxicity markers as functions of exposure duration as well as dose. The example used mortality for rats and mice, analyzed separately, from acute exposure to hydrogen sulfide (dose refers to airborne concentration of H(2)S). Statistical methods were applied to determine when data from different studies could be pooled. EC01, EC10, and EC50 (doses with response rates of 1, 10, and 50%) were estimated, with 95% confidence intervals, at durations of 5, 10, and 30 min, and 1, 2, 4, and 6 h. A single dose-duration response curve for mortality was fit to the rat data for exposures of 5 min, 10 min, 30 min, and 1h, using a logistic curve additive in log(dose) and log(duration). Separate fits of that model were required, however, at 2, 4, and 6h, due to an increasing impact of duration relative to concentration as duration increased. The curves for rats fit the data exceedingly well and exhibited a threshold-like response followed by a steep incline as concentration increased. There were fewer data for mice but the response pattern for mortality clearly differed from rats. This example demonstrates the feasibility of extending the concept of single-study benchmark doses to multiple-study dose-duration benchmarks, using U.S. EPA's program CatReg. Similar applications to long-term animal studies could be considered.

Air Pollutants↗

PubMed1998

Optimized dose distribution of a high dose rate vaginal cylinder.

PURPOSE: To present a comparison of optimized dose distributions for a set of high-dose-rate (HDR) vaginal cylinders calculated by a commercial treatment-planning system with benchmark calculations using Monte-Carlo-calculated dosimetry data. METHODS AND MATERIALS: Optimized dose distributions using both an isotropic and an anisotropic dose calculation model were obtained for a set of HDR vaginal cylinders. Mathematical optimization techniques available in the computer treatment-planning system were used to calculate dwell times and positions. These dose distributions were compared with benchmark calculations with TG43 formalism and using Monte-Carlo-calculated data. The same dwell times and positions were used for a quantitative comparison of dose calculated with three dose models. RESULTS: The isotropic dose calculation model can result in discrepancies as high as 50%. The anisotropic dose calculation model compared better with benchmark calculations. The differences were more significant at the apex of the vaginal cylinder, which is typically used as the prescription point. CONCLUSION: Dose calculation models available in a computer treatment-planning system must be evaluated carefully to ensure their correct application. It should also be noted that when optimized dose distribution at a distance from the cylinder surface is calculated using an accurate dose calculation model, the vaginal mucosa dose becomes significantly higher, and therefore should be carefully monitored.

Brachytherapy↗

PubMed2006

Simulating the effect of DNA polymerase mutations on transition-state energetics and fidelity: evaluating amino acid group contribution and allosteric coupling for ionized residues in human pol beta.

The control of the catalytic power and fidelity of DNA polymerases involves the complex combined effect of the protein residues, the Mg2+ ions, and the interaction between the DNA bases. In an attempt to advance the understanding of catalytic control, we analyze the effect of the protein residues, taking human DNA polymerase beta as a model system. Specifically, we examine the ability of different theoretical models to reproduce the effect of ionized residues on the transition state (TS) binding energy and the corresponding k(pol)/KD. We also explore the role of the Mg2+ ions in the binding and catalysis processes. The application of the microscopic linear response approximation (LRA) and the semimacroscopic PDLD/S-LRA methods to a benchmark of mutational studies produces a semiquantitative correlation and indicates that these methods can provide predictive power. However, pre-steady-state and steady-state kinetic studies currently available do not give a unique benchmark, owing principally to widely varying experimental conditions. We believe that a more uniform experimental benchmark is needed for further refinement of the theoretical models. The analysis of the correlation between the results obtained by a rigorous thermodynamic cycle and by simpler approximations indicates that the protein reorganization between the open, i.e., unbound, form and the closed form does not change the magnitude of the calculated mutational effects in a major way for the experimental data used in this study. The use of the PDLD/S-LRA group contributions allows us to construct energy-based correlation diagrams that can help toward understanding the coupling, i.e., transfer of information, between the base-binding and catalytic sites and to gain a deeper insight into the molecular basis of DNA replication fidelity. Our analysis suggests that the allosteric matrix obtained by subtracting the correlation matrix of the correct and incorrect base pairs should prove useful in exploring the information transfer occurring between the base-binding and catalytic sites. This type of treatment should be especially effective when coupled with structural studies of polymerase-DNA-base mispair ternary complexes and studies using polymerase double mutants. We discuss the potential of direct calculations of binding energy of the TS in a rational design of TS analogues and in drug design.

Allosteric Regulation↗

PubMed2004

Brief survey of EPA standard-setting and health assessment.

The Environmental Protection Agency (EPA) promulgates standards for air pollutants and drinking water contaminants, as part of its mandate to protect public health and welfare. The Agency also assesses the health risks associated with hundreds of chemical substances, often developing quantitative toxicity and cancer potency benchmarks. This article compares EPA standards and benchmark values to those of other countries and other agencies. This includes the national ambient air quality standards (NAAQS), the national primary drinking water regulations (NPDWR), and benchmark values from the Integrated Risk Information System (IRIS). Results show that the NAAQS are generally comparable to or less strict than the air quality standards of other countries and international organizations. The NPDWR tend to be less strict than the water quality standards of other countries for inorganic chemicals, and they are more strict about as often as they are less strict for organic chemicals. Reference values for toxicity and cancer potency derived in EPA health assessments posted in the IRIS database are less stringentthan those of other agencies about as often as they are more stringent, and they are often identical. Revisions to these values more often than not made them less stringent. These results suggest that EPA's standards and quantitative health assessments are not out of line with those of other agencies and other countries.

Carcinogenicity Tests↗

PubMed2005

Through the channel and around the channel: Validating and comparing microscopic approaches for the evaluation of free energy profiles for ion penetration through ion channels.

Microscopic calculations of free energy profiles for ion transport through biological ion channels present a very serious challenge to modern simulation approaches. The main problem is due to the major convergence problems associated with the heterogeneous landscape of the electrostatic environment in ion channels and with the need to evaluate the profile associated with the transfer of the ion from bulk water to the channel environment. This problem is compounded by the lack of reliable and relevant benchmarks that can discriminate between alternative approaches. The present study is aimed at reducing the above problems by defining benchmarks that are directly relevant to ion channels and can also give converging results. This is done by constructing a series of models of a truncated gramicidin channel with different numbers of water molecules and by comparing the profiles for going around the channel and through the channel. These discriminating models are then used to validate and compare the adiabatic charging free energy perturbation (FEP) approach combined with an umbrella sampling approach (Warshel, A. J. Phys. Chem. 1982, 86, 2218) and the potential of mean force (PMF) approach used frequently in studies of ion channels. It is found that both approaches work quite well until one moves to the case of the fully solvated channel. In this limit, the PMF approach may give different results for the overall work of going through the channel and around the channel, while the FEP approach gives physically consistent results. The present benchmark also indicates that the weighted histogram analysis method (WHAM) approach does not offer a significant advantage over earlier approaches at least as much as studies of ion channels are concerned. Finally, it is concluded that the FEP approach may be more useful in evaluating the overall barrier for moving ions from water to ion channels and that in some cases it might be beneficial to use the FEP approach for selective points along the channel and then to connect these points by PMF calculations.

Algorithms↗

PubMed2005

A ballistic model of choice response time.

Almost all models of response time (RT) use a stochastic accumulation process. To account for the benchmark RT phenomena, researchers have found it necessary to include between-trial variability in the starting point and/or the rate of accumulation, both in linear (R. Ratcliff & J. N. Rouder, 1998) and nonlinear (M. Usher & J. L. McClelland, 2001) models. The authors show that a ballistic (deterministic within-trial) model using a simplified version of M. Usher and J. L. McClelland's (2001) nonlinear accumulation process with between-trial variability in accumulation rate and starting point is capable of accounting for the benchmark behavioral phenomena. The authors successfully fit their model to R. Ratcliff and J. N. Rouder's (1998) data, which exhibit many of the benchmark phenomena.

Choice Behavior↗

PubMed2026

Prediction of human missense variant effects from functional evidence.

Prediction of missense variant effects remains a critical bottleneck in both research and diagnostic genetics. Current predictors typically rely on clinical outcomes or population patterns rather than direct measures of functional impact, leading to limited generalizability and data circularity. Here we present FuncVEP, a family of variant effect predictors trained on diverse functional data to predict the functional impact of missense variants. FuncVEP generalizes across datasets and outperforms 48 existing predictors across a wide range of benchmarks, improving accuracy from 78.8% to 84.6% on functional benchmarks and from 90.1% to 92.4% on clinical benchmarks. From a discovery perspective, we identified 210 new gene-phenotype associations involving 494 genes linked to inborn errors of immunity in the UK Biobank and the Mount Sinai Million Health Discoveries Program. FuncVEP substantially improved the discovery rate relative to state-of-the-art predictors. Overall, FuncVEP provides a robust, scalable solution for variant interpretation, advancing both diagnostic precision and gene discovery.

Humans↗

PubMed2005

The extrapolation of one-electron basis sets in electronic structure calculations: how it should work and how it can be made to work.

We consider the extrapolation of the one-electron basis to the basis set limit in the context of coupled cluster calculations. We produce extrapolation coefficients that produce much more accurate results than previous extrapolation forms. These are determined by fitting to accurate benchmark results. For coupled cluster singles doubles energies, we take our benchmark results from the work of Klopper that explicitly includes the interelectronic distance. For the perturbative triples energies, our benchmark results are obtained from large even-tempered basis set calculations.

Journal Article↗

PubMed2003

Use of audit tools to evaluate the efficacy of cleaning systems in hospitals.

BACKGROUND: Recent publicity has highlighted both the inadequacies of hospital cleaning and high levels of methicillin-resistant Staphylococcus aureus infections in UK hospitals. "Standards for Environmental Cleanliness" (SEC) was a checklist developed in April 1999 by the Infection Control Nurses Association and the Association of Domestic Managers to evaluate cleaning services regardless of who is the provider. More recently, the National Health Service plan (July 2000) was an attempt to generate a rapid improvement in the cleanliness and tidiness of hospitals via a National Health Service patient environment audit (PEA). On the basis of models used in the food industry to manage cleaning practices cost-effectively, a risk-based audit checklist incorporating rapid hygiene monitoring was developed to assess the adequacy of cleaning programs and standards in hospitals. This checklist (Audit for Cleaning Efficacy, or ACE) as well as the SEC and PEA approaches were applied at 4 hospitals, and environmental microbial surface counts were compared. SEC and PEA rely on visual assessment, whereas the ACE approach is more comprehensive and included more specific questions relating to the management and monitoring of cleaning as well as standards on the basis of rapid hygiene monitoring. METHODS: Two wards in each of the 4 hospitals were visited on 3 separate occasions immediately after cleaning was completed. Visual assessment, adenosine triphosphate bioluminescence, and microbiologic sampling of selected environmental sites were performed to evaluate the effectiveness of cleaning. The 3 audits were completed during the final hospital visit. RESULTS: Visual assessment indicated that 90% of sites were satisfactory, whereas adenosine triphosphate bioluminescence showed that 100% and microbiologic sampling showed that 90% of sites did not meet benchmark values. There was no significant difference between the SEC and PEA audits (P =.311), which used visual assessment, and the results suggest that they both are similar in passing surfaces that have microbiologic benchmark values that are too high. However, the ACE audit showed a significant difference (P = <.001) in results compared with the SEC and the PEA audits and did not pass surfaces with microbiologic benchmark values that were too high. The ACE audit, which incorporates rapid hygiene testing, showed a much stronger association with the microbial counts; this was not apparent with the SEC and the PEA audits. CONCLUSION: The data suggest that visual assessment is a poor indicator of cleaning efficacy and that the ACE audit gives a better assessment of cleaning programs compared with the other 2 audit methods in relation to microbial surface counts. It is recommended that hospital cleaning regimes be designed to ensure that surfaces are cleaned adequately and that efficacy is assessed with use of internal auditing and rapid hygiene testing.

Adenosine Triphosphate↗

PubMed2004

Automated structure prediction of weakly homologous proteins on a genomic scale.

We have developed TASSER, a hierarchical approach to protein structure prediction that consists of template identification by threading, followed by tertiary structure assembly via the rearrangement of continuous template fragments guided by an optimized C(alpha) and side-chain-based potential driven by threading-based, predicted tertiary restraints. TASSER was applied to a comprehensive benchmark set of 1,489 medium-sized proteins in the Protein Data Bank. With homologues excluded, in 927 cases, the templates identified by our threading algorithm PROSPECTOR_3 have a rms deviation from native <6.5 A with approximately 80% alignment coverage. After template reassembly, this number increases to 1,172. This shows significant and systematic improvement of the final models with respect to the initial template alignments. Furthermore, significant improvements in loop modeling are demonstrated. We then apply TASSER to the 1,360 medium-sized ORFs in the Escherichia coli genome; approximately 920 can be predicted with high accuracy based on confidence criteria established in the Protein Data Bank benchmark. These results from our unprecedented comprehensive folding benchmark on all protein categories provide a reliable basis for the application of TASSER to structural genomics, especially to proteins of low sequence identity to solved protein structures.

Algorithms↗

PubMed2003

Influence of seat properties on occupant dynamics in severe rear crashes.

Seat performance in retaining an occupant, transferring energy, and controlling neck responses is often questioned after severe rear crashes when fatal or disabling injury occur. It is argued that a stiffer seat would have improved occupant kinematics. However, there are many factors in occupant interactions with the seat. This study evaluates four different seat types in 26 and 32 mph (42 and 51 km/h), rear crash delta Vs. Two seats were yielding with k = 20 kN/m occupant load per displacement. One represented a 1970s yielding seat with j = 3.4 degrees /kN frame rotation per occupant load, and 3 kN maximum load (660 Nm moment), and the other a high retention seat phased into production since 1997 with j = 1.4 degrees /kN, and 10 kN maximum load (2200 Nm). Two seats were stiff with k = 40 kN/m. One represented a 1990s foreign benchmark with j = 1.8 degrees /kN and a 7.7 kN maximum load (1700 Nm), and the other an all belts to seat (ABTS) with j = 1.0 degrees /kN and 20 kN maximum load (4400 Nm). The crash was a constant acceleration of 11.8 g, or 14.5 g for 100 ms. Occupant interactions with the seat were modeled using a torso mass, flexible neck and head mass. By analysis of the equations of motion, the initial change in seatback angle (Deltatheta) is proportional to jk(y - x), the product jk and the differential motion between the vehicle (seat cushion) and occupant. The transition from 1970s-80s yielding seats to stronger seats of the 1990s involved an increase in k stiffness; however, the jk property did not change as frame structures became stronger. The yielding seats of the 1970s had jk = 68 degrees /m, while the stiff foreign benchmark seat had jk = 72 degrees /m. The foreign benchmark rotated about the same as the 1970s seat up to 50 ms in the severe rear crashes. While it was substantially stronger, it produced higher loads on the occupant, and the higher loads increased seatback rotations and neck responses. The ABTS seat had the lowest rotations but also caused high neck responses because of the greater loads on the torso. Neck displacement (d) is initially proportional to (k/m(T)) integral integral y, seat stiffness times the second integral of vehicle displacement divided by torso mass. As seat stiffness increases, head-torso acceleration, velocity, and neck displacement increase. This study shows that the jk seat property determines the initial seatback rotation in rear crashes. If a stronger seat has a higher stiffness, it rotates at higher loads on the occupant, reducing the overall benefit of the stronger frame, while increasing neck responses related to whiplash or neck extension prior to subsequent impacts. The aim of seat designs should be to reduce jk, provide pocketing of the pelvis, and give head-neck support for the best protection in severe rear crashes. For low-speed crashes, a low k is important to reduce early neck responses related to whiplash.

Acceleration↗

PubMed1993

Monte Carlo calculation of kerma to a point in the vicinity of media interfaces.

Recent work has demonstrated that Monte Carlo photon transport (MCPT) simulation is a practical and accurate dosimetry tool for characterizing brachytherapy dose distributions in the presence of complex three-dimensional (3D) geometries. Because of the steep dose gradients encountered in this application, the choice of estimation algorithm, i.e., the algorithm for extracting dose estimates from individual photon histories, is critical to ensure adequate spatial resolution and computational efficiency. While conventional MCPT dose estimation techniques are sufficiently accurate in many cases, they inevitably fail when the point of interest is located near media interfaces, since they require the point of interest to be contained by a region of homogeneous medium of several millimetres in diameter. This severely limits the domain of application of MCPT in brachytherapy dosimetry and prevents dose estimation at points such as those immediately behind lead shields contained in gynaecological applicators, or simulation of small intricately structured detector response. We have successfully adapted to an MCPT code a neutron-flux estimator, the once-more-collided flux estimator (OMCFE), which overcomes this problem. OMCFE is derived from the familiar next-flight estimation algorithm, and, through the use of a photon trajectory resampling scheme, does not require homogeneous media at the point of interest. It is therefore a true point-dose estimator. OMCFE was tested against a variety of one- and three-dimensional benchmark problems. These benchmarks included mono-energetic point sources of 30 and 600 keV, as well as a 3M model 6702 125I seed, a 137Cs tube, and an 192Ir seed in geometries approximating common brachytherapy dosimetric problems. OMCFE agrees closely with conventional MCPT dose estimation algorithms in solving these benchmark problems, while in many cases being much more efficient. It therefore makes possible brachytherapy dose estimation with high spatial resolution, in contrast to the larger than 1 mm spatial resolution achievable using conventional MCPT estimation algorithms. It should be noted that the method presented in this paper is only valid when charged-particle equilibrium is assumed.

Electrons↗

PubMed2026

Agentomics: an agentic system that autonomously develops novel state-of-the-art solutions for biomedical machine learning tasks.

MOTIVATION: Extracting knowledge from biomedical data is crucial for advancing our understanding of biological systems and developing novel therapeutics. The quantity, quality, and resolution of biomedical data constantly evolves, requiring the automation of biomedical machine learning (ML). Existing Automated ML tools lack flexibility, while large language models (LLMs) struggle to consistently deliver reproducible machine learning codebases, and existing LLM Agent-powered solutions lag behind human-engineered ML models. RESULTS: Here, we introduce Agentomics, an autonomous LLM-powered agentic system for end-to-end ML experimentation. Given a biomedical dataset, Agentomics implements various ML modeling strategies, and produces a ready-to-use ML model. Agentomics introduces strict validation checkpoints for standard ML development steps, allowing gradual development on top of working code with defined interfaces and validated artifacts. Further, it offers native support for biomedical foundation models that can be leveraged during experimentation. The generic nature of Agentomics allows the user to create ML solutions for a large variety of datasets and use various LLMs. We evaluate Agentomics across 20 datasets from the domains of Protein Engineering, Drug Discovery, and Regulatory Genomics. When benchmarked against other agentic systems, Agentomics outperformed them in all tested domains. When benchmarked against human expert solutions, Agentomics generated novel state-of-the-art models for 11/20 established benchmark datasets. AVAILABILITY AND IMPLEMENTATION: Agentomics is implemented in Python. Source code and documentation are freely available at: https://github.com/BioGeMT/Agentomics-ML.

Machine Learning↗