Your browser doesn't support javascript.
loading
Mostrar: 20 | 50 | 100
Resultados 1 - 20 de 127.412
Filtrar
1.
Braz. j. biol ; 84: e245592, 2024. tab, graf
Artigo em Inglês | LILACS, VETINDEX | ID: biblio-1355866

RESUMO

Abstract In recent years, the development of high-throughput technologies for obtaining sequence data leveraged the possibility of analysis of protein data in silico. However, when it comes to viral polyprotein interaction studies, there is a gap in the representation of those proteins, given their size and length. The prepare for studies using state-of-the-art techniques such as Machine Learning, a good representation of such proteins is a must. We present an alternative to this problem, implementing a fragmentation and modeling protocol to prepare those polyproteins in the form of peptide fragments. Such procedure is made by several scripts, implemented together on the workflow we call PolyPRep, a tool written in Python script and available in GitHub. This software is freely available only for noncommercial users.


Resumo Nos últimos anos, o desenvolvimento de tecnologias de alto rendimento para obtenção de dados sequenciais potencializou a possibilidade de análise de dados proteicos in silico. No entanto, quando se trata de estudos de interação de poliproteínas virais, existe uma lacuna na representação dessas proteínas, devido ao seu tamanho e comprimento. Para estudos utilizando técnicas de ponta como o Aprendizado de Máquina, uma boa representação dessas proteínas é imprescindível. Apresentamos uma alternativa para este problema, implementando um protocolo de fragmentação e modelagem para preparar essas poliproteínas na forma de fragmentos de peptídeos. Tal procedimento é feito por diversos scripts, implementados em conjunto no workflow que chamamos de PolyPRep, uma ferramenta escrita em script Python e disponível no GitHub. Este software está disponível gratuitamente apenas para usuários não comerciais.


Assuntos
Protease de HIV , Poliproteínas , Software , Simulação de Acoplamento Molecular
2.
Methods Mol Biol ; 2713: 519-541, 2024.
Artigo em Inglês | MEDLINE | ID: mdl-37639145

RESUMO

Cell morphology and motility drive the cellular capabilities to interact with the environment. For example, microglia, the longest known tissue-resident macrophages, show a highly branched process tree with which they continuously scan their environment. Computational image analysis allows to quantify morphology and/or motility from images of tissue-resident macrophages. Here, I describe a step-by-step protocol for analyzing the morphology (and motility) of macrophages with our recently described, freely available software MotiQ, which provides a broad band of parameters and thereby serves as a versatile tool for studies of morphology and motility.


Assuntos
Macrófagos , Microglia , Processamento de Imagem Assistida por Computador , Software , Árvores
3.
Methods Mol Biol ; 2713: 505-518, 2024.
Artigo em Inglês | MEDLINE | ID: mdl-37639144

RESUMO

Using the open-source image analysis software CellProfiler to automatically quantify antibody-stained or fluorescently labeled macrophages in situ provides accurate and reproducible cell counts. It is a vastly enhanced alternative method to both manual cell counting and estimation of cell marker expression based on fluorescence intensity. Quantification of tissue-resident macrophages acquired on widefield or confocal microscopes can be batch processed using our pipeline to produce data within minutes.


Assuntos
Anticorpos , Macrófagos , Contagem de Células , Processamento de Imagem Assistida por Computador , Software
4.
Methods Mol Biol ; 2716: 181-202, 2024.
Artigo em Inglês | MEDLINE | ID: mdl-37702940

RESUMO

The high-performance computing (HPC) platform for large-scale drug discovery simulation demands significant investment in speciality hardware, maintenance, resource management, and running costs. The rapid growth in computing hardware has made it possible to provide cost-effective, robust, secure, and scalable alternatives to the on-premise (on-prem) HPC via Cloud, Fog, and Edge computing. It has enabled recent state-of-the-art machine learning (ML) and artificial intelligence (AI)-based tools for drug discovery, such as BERT, BARD, AlphaFold2, and GPT. This chapter attempts to overview types of software architectures for developing scientific software or application with deployment agnostic (on-prem to cloud and hybrid) use cases. Furthermore, the chapter aims to outline how the innovation is disrupting the orthodox mindset of monolithic software running on on-prem HPC and provide the paradigm shift landscape to microservices driven application programming (API) and message parsing interface (MPI)-based scientific computing across the distributed, high-available infrastructure. This is coupled with agile DevOps, and good coding practices, low code and no-code application development frameworks for cost-efficient, secure, automated, and robust scientific application life cycle management.


Assuntos
Inteligência Artificial , Computação em Nuvem , Algoritmos , Descoberta de Drogas , Software
5.
Methods Mol Biol ; 2716: 51-99, 2024.
Artigo em Inglês | MEDLINE | ID: mdl-37702936

RESUMO

Modeling and simulation (M&S), including in silico (clinical) trials, helps accelerate drug research and development and reduce costs and have coined the term "model-informed drug development (MIDD)." Data-driven, inferential approaches are now becoming increasingly complemented by emerging complex physiologically and knowledge-based disease (and drug) models, but differ in setup, bottlenecks, data requirements, and applications (also reminiscent of the different scientific communities they arose from). At the same time, and within the MIDD landscape, regulators and drug developers start to embrace in silico trials as a potential tool to refine, reduce, and ultimately replace clinical trials. Effectively, silos between the historically distinct modeling approaches start to break down. Widespread adoption of in silico trials still needs more collaboration between different stakeholders and established precedence use cases in key applications, which is currently impeded by a shattered collection of tools and practices. In order to address these key challenges, efforts to establish best practice workflows need to be undertaken and new collaborative M&S tools devised, and an attempt to provide a coherent set of solutions is provided in this chapter. First, a dedicated workflow for in silico clinical trial (development) life cycle is provided, which takes up general ideas from the systems biology and quantitative systems pharmacology space and which implements specific steps toward regulatory qualification. Then, key characteristics of an in silico trial software platform implementation are given on the example of jinko.ai (nova's end-to-end in silico clinical trial platform). Considering these enabling scientific and technological advances, future applications of in silico trials to refine, reduce, and replace clinical research are indicated, ranging from synthetic control strategies and digital twins, which overall shows promise to begin a new era of more efficient drug development.


Assuntos
Desenvolvimento de Medicamentos , Bases de Conhecimento , Simulação por Computador , Memória , Software
6.
Methods Enzymol ; 688: 115-143, 2023.
Artigo em Inglês | MEDLINE | ID: mdl-37748824

RESUMO

Molecular-dynamics (MD) simulations have contributed substantially to our understanding of protein structure and dynamics, yielding insights into many biological processes including protein folding, drug binding, and mechanisms of protein-protein interactions. Much of what is known about protein structure comes from macromolecular crystallography (MX) experiments. MD simulations of protein crystals are useful in the study of MX because the simulations can be analyzed to calculate almost any crystallographic observable of interest, from atomic coordinates to structure factors and densities, B-factors, multiple conformations and their populations/occupancies, and diffuse scattering intensities. Computing resources and software to support crystalline MD simulations are now readily available to many researchers studying protein structure and dynamics and who may be interested in advanced interpretation of MX data, including diffuse scattering. In this work, we outline methods of analyzing MD simulations of protein crystals and provide accompanying Jupyter notebooks as practical resources for researchers wishing to perform similar analyses on their own systems of interest.


Assuntos
Simulação de Dinâmica Molecular , Dobramento de Proteína , Cristalografia , Substâncias Macromoleculares , Software
7.
J Neuroeng Rehabil ; 20(1): 125, 2023 09 26.
Artigo em Inglês | MEDLINE | ID: mdl-37749627

RESUMO

BACKGROUND: 'Perturbation-based balance training' (PBBT) is a training method that was developed to improve balance reactive responses to unexpected balance loss. This training method is more effective in reducing fall rates than traditional balance training methods. Many PBBTs are performed during standing or treadmill walking which targeted specifically step reactive responses, we however, aimed to develop and build a mechatronic system that can provide unexpected perturbation during elliptical walking the Elliptical Perturbation System (the EPES system), with the aim of improving specifically the trunk and upper limbs balance reactive control. METHODS: This paper describes the development, and building of the EPES system, using a stationary Elliptical Exercise device, which allows training of trunk and upper limbs balance reactive responses in older adults. RESULTS: The EPES system provides 3-dimensional small, controlled, and unpredictable sudden perturbations during stationary elliptical walking. We developed software that can identify a trainee's trunk and arms reactive balance responses using a stereo camera. After identifying an effective trunk and arms reactive balance response, the software controls the EPES system motors to return the system to its horizontal baseline position after the perturbation. The system thus provides closed-loop feedback for a person's counterbalancing trunk and arm responses, helping to implement implicit motor learning for the trainee. The pilot results show that the EPES software can successfully identify balance reactive responses among participants who are exposed to a sudden unexpected perturbation during elliptical walking on the EPES system. CONCLUSIONS: EPES trigger reactive balance responses involving counter-rotation action of body segments and simultaneously evoke arms, and trunk reactive response, thus reactive training effects should be expected.


Assuntos
Exercício Físico , Caminhada , Humanos , Idoso , Extremidade Superior , Rotação , Software
8.
Sci Rep ; 13(1): 15817, 2023 09 22.
Artigo em Inglês | MEDLINE | ID: mdl-37740101

RESUMO

Rapid urbanization, population growth, agricultural practices, and industrial activities have led to widespread groundwater contamination. This study evaluated heavy metal contamination in residential drinking water in Shiraz, Iran (2021). The analysis involved 80 groundwater samples collected across wet and dry seasons. Water quality was comprehensively assessed using several indices, including the heavy metals evaluation index (HEI), heavy metal pollution index (HPI), contamination degree (CD), and metal index (MI). Carcinogenic and non-carcinogenic risk assessments were conducted using deterministic and probabilistic approaches for exposed populations. In the non-carcinogenic risk assessment, the chronic daily intake (CDI), hazard quotient (HQ), and hazard index (HI) are employed. The precision of risk assessment was bolstered through the utilization of Monte Carlo simulation, executed using the R software platform. Based on the results, in both wet and dry seasons, Zinc (Zn) consistently demonstrates the highest mean concentration, followed by Manganese (Mn) and Chromium (Cr). During the wet and dry seasons, 25% and 40% of the regions exhibited high CD, respectively. According to non-carcinogenic risk assessment, Cr presents the highest CDI and HQ in children and adults, followed by Mn, As and HI values, indicating elevated risk for children. The highest carcinogenic risk was for Cr in adults, while the lowest was for Cd in children. The sensitivity analysis found that heavy metal concentration and ingestion rate significantly impact both carcinogenic and non-carcinogenic risks. These findings provide critical insights for shaping policy and allocating resources towards effectively managing heavy metal contamination in residential drinking water.


Assuntos
Água Potável , Metais Pesados , Adulto , Criança , Humanos , Método de Monte Carlo , Qualidade da Água , Software , Cromo , Manganês , Carcinógenos , Carcinogênese
9.
J Vis Exp ; (199)2023 09 08.
Artigo em Inglês | MEDLINE | ID: mdl-37747186

RESUMO

Protein quantitation is an essential procedure in life sciences research. Amongst several other methods, the Bradford assay is one of the most used. Because of its widespread, the limitations and advantages of the Bradford assay have been exhaustively reported, including several modifications of the original method to improve its performance. One of the alterations of the original method is the use of a smartphone camera as an analytical instrument. Taking advantage of the three forms of the Coomassie Brilliant Blue dye that exist in the conditions of the Bradford assay, this paper describes how to accurately quantify protein in samples using color data extracted from a single picture of a microplate. After performing the assay in a microplate, a picture is taken using a smartphone camera, and RGB color data is extracted from the picture using a free and open-source image analysis software application. Then, the ratio of blue to green intensity (in the RGB scale) of samples with unknown concentrations of protein is used to calculate the protein content based on a standard curve. No significant difference is observed between values calculated using RGB color data and those calculated using conventional absorbance data.


Assuntos
Bioensaio , Smartphone , Processamento de Imagem Assistida por Computador , Software
10.
Proc Natl Acad Sci U S A ; 120(40): e2310488120, 2023 Oct 03.
Artigo em Inglês | MEDLINE | ID: mdl-37748054

RESUMO

Cognitive scientists treat verification as a computation in which descriptions that match the relevant situation are true, but otherwise false. The claim is controversial: The logician Gödel and the physicist Penrose have argued that human verifications are not computable. In contrast, the theory of mental models treats verification as computable, but the two truth values of standard logics, true and false, as insufficient. Three online experiments (n = 208) examined participants' verifications of disjunctive assertions about a location of an individual or a journey, such as: 'You arrived at Exeter or Perth'. The results showed that their verifications depended on observation of a match with one of the locations but also on the status of other locations (Experiment 1). Likewise, when they reached one destination and the alternative one was impossible, their use of the truth value: could be true and could be false increased (Experiment 2). And, when they reached one destination and the only alternative one was possible, they used the truth value, true and it couldn't have been false, and when the alternative one was impossible, they used the truth value: true but it could have been false (Experiment 3). These truth values and those for falsity embody counterfactuals. We implemented a computer program that constructs models of disjunctions, represents possible destinations, and verifies the disjunctions using the truth values in our experiments. Whether an awareness of a verification's outcome is computable remains an open question.


Assuntos
Médicos , Humanos , Software
11.
Proc Natl Acad Sci U S A ; 120(40): e2311005120, 2023 Oct 03.
Artigo em Inglês | MEDLINE | ID: mdl-37748055

RESUMO

Over the last decade, the United States has seen increasing antidemocratic rhetoric by political leaders. Yet, prior work suggests that such norm-violating rhetoric does not undermine support for democracy as a system of government. We argue that, while that may be true, such rhetoric does vitiate support for specific democratic principles. We test this theory by extending prior work to assess the effects of Trump's norm-violating rhetoric on general support for democracy as well as for the principles of participatory inclusiveness, contestation, the rule of law, and political equality. We find that Trump's rhetoric does not alter attitudes toward democracy as a preferred system but does reduce support for inclusiveness and equality among his supporters. Our findings suggest that elite rhetoric can undermine basic principles of American democracy.


Assuntos
Governo , Software
12.
Methods Enzymol ; 688: 1-42, 2023.
Artigo em Inglês | MEDLINE | ID: mdl-37748823

RESUMO

A long-standing goal in X-ray crystallography has been to extract information about the collective motions of proteins from diffuse scattering: the weak, textured signal that is found in the background of diffraction images. In the past few years, the field of macromolecular diffuse scattering has seen dramatic progress, and many of the past challenges in measurement and interpretation are now considered tractable. However, the concept of diffuse scattering is still new to many researchers, and a general set of procedures needed to collect a high-quality dataset has never been described in detail. Here, we provide the first guidelines for performing diffuse scattering experiments, which can be performed at any macromolecular crystallography beamline that supports room-temperature studies with a direct detector. We begin with a brief introduction to the theory of diffuse scattering and then walk the reader through the decision-making processes involved in preparing for and conducting a successful diffuse scattering experiment. Finally, we define quality metrics and describe ways to assess data quality both at the beamline and at home. Data obtained in this way can be processed independently by crystallographic software and diffuse scattering software to produce both a crystal structure, which represents the average atomic coordinates, and a three-dimensional diffuse scattering map that can then be interpreted in terms of models for protein motions.


Assuntos
Software , Síncrotrons , Coleta de Dados , Cristalografia por Raios X , Movimento (Física)
13.
Methods Enzymol ; 688: 195-222, 2023.
Artigo em Inglês | MEDLINE | ID: mdl-37748827

RESUMO

This chapter discusses the use of diffraction simulators to improve experimental outcomes in macromolecular crystallography, in particular for future experiments aimed at diffuse scattering. Consequential decisions for upcoming data collection include the selection of either a synchrotron or free electron laser X-ray source, rotation geometry or serial crystallography, and fiber-coupled area detector technology vs. pixel-array detectors. The hope is that simulators will provide insights to make these choices with greater confidence. Simulation software, especially those packages focused on physics-based calculation of the diffraction, can help to predict the location, size, shape, and profile of Bragg spots and diffuse patterns in terms of an underlying physical model, including assumptions about the crystal's mosaic structure, and therefore can point to potential issues with data analysis in the early planning stages. Also, once the data are collected, simulation may offer a pathway to improve the measurement of diffraction, especially with weak data, and might help to treat problematic cases such as overlapping patterns.


Assuntos
Análise de Dados , Software , Simulação por Computador , Cristalografia , Substâncias Macromoleculares
14.
J Med Life ; 16(6): 862-867, 2023 Jun.
Artigo em Inglês | MEDLINE | ID: mdl-37675166

RESUMO

High-quality and detailed CT scan images are crucial for accurate diagnosis. Factors such as image noise and slice thickness affect image quality. This study aimed to determine the optimal slice thickness that minimized image noise while maintaining sufficient diagnostic information using the single-source computed tomography head protocol. Single-source CT images were examined using the Linux Operating system Ge Revolution 64-slice CT scanner, and a combination of statical analysis and DICOM CT image analysis was employed. The single-source energy head CT protocol was used to investigate the effect of slice thickness on noise and visibility in images. Different values of slice thickness 0.625, 1.25, 2.5, 3.75, 5, 7.5, and 10 were prepared, and then quantitative analysis was performed. Thinner slice thickness decreased image noise, increased visibility, and improved detection. Therefore, the balance between changing the thickness of the slice with the diagnostic content and image noise must be considered. Maximum slice thickness enhances CT image detail and structure despite more noise. Based on the results, a slice thickness of 1.25mm was identified as the optimal choice for reducing image noise and achieving better and more accurate detection using the single-source computed tomography head protocol. The study revealed that image noise tends to increase with greater slice thickness according to the Linux operating system. These findings can serve as a valuable guide for quality control methods in CT centers, emphasizing the need to determine the appropriate slice thickness to ensure an accurate diagnosis.


Assuntos
Processamento de Imagem Assistida por Computador , Tomografia Computadorizada por Raios X , Humanos , Controle de Qualidade , Software
15.
J Cancer Res Ther ; 19(4): 988-994, 2023 Aug.
Artigo em Inglês | MEDLINE | ID: mdl-37675727

RESUMO

Context: The present experimental models of cystic diseases are not adequate and require further investigation. Aim: In this study, a new way of producing a tissue-mimicking model of cysts and cystic neoplasms was evaluated. Settings and Design: To simulate cysts and cystic neoplasms, ex vivo rabbit normal bladders and VX2-implanted tumor bladders were produced, fixed, and embedded in agarose gel. Methods and Materials: The samples were classified into four groups based on tumor features and the maximal transverse diameter of the rabbit bladder, which were assessed using computer tomography (CT) imaging and statistically analyzed. Statistical Analysis Used: Statistical analysis was performed using Statistical Package for the Social Sciences (SPSS) software. The t-test was used for analyzing enumeration data. Results: Twenty-one rabbit bladders (21/24) were successfully removed and prepped for this experiment, comprising eleven normal bladders (11/24) and ten implanted with VX2 tumors (10/24). The gelling ingredient used to form the visualization and fixation matrix was agarose at a concentration of 4 g/200 mL. The temperature of the agarose solution was kept constant at 40-45°C, which is the optimal temperature range for ex vivo normal bladder and implanted VX2 tumor bladder insertion. The average time required to embed and fix the bladders in agarose gel was 45.0 ± 5.2 minutes per instance. The gel-fixing matrix's strength and light transmittance were enough for building the models. Conclusion: We created an experimental tissue-mimicking model of cysts and cystic neoplasms with stable physicochemical features, a safe manufacturing method, and high repeatability. These models may be used to assist with cystic lesion diagnosis and treatment techniques.


Assuntos
Cistos , Neoplasias Císticas, Mucinosas e Serosas , Neoplasias da Bexiga Urinária , Animais , Coelhos , Sefarose , Cistos/diagnóstico por imagem , Software
16.
BMC Bioinformatics ; 24(1): 336, 2023 Sep 11.
Artigo em Inglês | MEDLINE | ID: mdl-37697267

RESUMO

BACKGROUND: Residue Interaction Networks (RINs) map the crystallographic description of a protein into a graph, where amino acids are represented as nodes and non-covalent bonds as edges. Determination and visualization of a protein as a RIN provides insights on the topological properties (and hence their related biological functions) of large proteins without dealing with the full complexity of the three-dimensional description, and hence it represents an invaluable tool of modern bioinformatics. RESULTS: We present RINmaker, a fast, flexible, and powerful tool for determining and visualizing RINs that include all standard non-covalent interactions. RINmaker is offered as a cross-platform and open source software that can be used either as a command-line tool or through a web application or a web API service. We benchmark its efficiency against the main alternatives and provide explicit tests to show its performance and its correctness. CONCLUSIONS: RINmaker is designed to be fully customizable, from a simple and handy support for experimental research to a sophisticated computational tool that can be embedded into a large computational pipeline. Hence, it paves the way to bridge the gap between data-driven/machine learning approaches and numerical simulations of simple, physically motivated, models.


Assuntos
Aminoácidos , Benchmarking , Biologia Computacional , Aprendizado de Máquina , Software
17.
BMC Med Educ ; 23(1): 659, 2023 Sep 11.
Artigo em Inglês | MEDLINE | ID: mdl-37697275

RESUMO

BACKGROUND: Automated Item Generation (AIG) uses computer software to create multiple items from a single question model. There is currently a lack of data looking at whether item variants to a single question result in differences in student performance or human-derived standard setting. The purpose of this study was to use 50 Multiple Choice Questions (MCQs) as models to create four distinct tests which would be standard set and given to final year UK medical students, and then to compare the performance and standard setting data for each. METHODS: Pre-existing questions from the UK Medical Schools Council (MSC) Assessment Alliance item bank, created using traditional item writing techniques, were used to generate four 'isomorphic' 50-item MCQ tests using AIG software. Isomorphic questions use the same question template with minor alterations to test the same learning outcome. All UK medical schools were invited to deliver one of the four papers as an online formative assessment for their final year students. Each test was standard set using a modified Angoff method. Thematic analysis was conducted for item variants with high and low levels of variance in facility (for student performance) and average scores (for standard setting). RESULTS: Two thousand two hundred eighteen students from 12 UK medical schools participated, with each school using one of the four papers. The average facility of the four papers ranged from 0.55-0.61, and the cut score ranged from 0.58-0.61. Twenty item models had a facility difference > 0.15 and 10 item models had a difference in standard setting of > 0.1. Variation in parameters that could alter clinical reasoning strategies had the greatest impact on item facility. CONCLUSIONS: Item facility varied to a greater extent than the standard set. This difference may relate to variants causing greater disruption of clinical reasoning strategies in novice learners compared to experts, but is confounded by the possibility that the performance differences may be explained at school level and therefore warrants further study.


Assuntos
Raciocínio Clínico , Estudantes de Medicina , Humanos , Aprendizagem , Faculdades de Medicina , Software
18.
PLoS Comput Biol ; 19(9): e1011477, 2023 09.
Artigo em Inglês | MEDLINE | ID: mdl-37669275

RESUMO

Here, we introduce Trackplot, a Python package for generating publication-quality visualization by a programmable and interactive web-based approach. Compared to the existing versions of programs generating sashimi plots, Trackplot offers a versatile platform for visually interpreting genomic data from a wide variety of sources, including gene annotation with functional domain mapping, isoform expression, isoform structures identified by scRNA-seq and long-read sequencing, as well as chromatin accessibility and architecture without any preprocessing, and also offers a broad degree of flexibility for formats of output files that satisfy the requirements of major journals. The Trackplot package is an open-source software which is freely available on Bioconda (https://anaconda.org/bioconda/trackplot), Docker (https://hub.docker.com/r/ygidtu/trackplot), PyPI (https://pypi.org/project/trackplot/) and GitHub (https://github.com/ygidtu/trackplot), and a built-in web server for local deployment is also provided.


Assuntos
Cromatina , Genômica , Anotação de Sequência Molecular , Registros , Software
19.
PLoS Comput Biol ; 19(9): e1011454, 2023 09.
Artigo em Inglês | MEDLINE | ID: mdl-37669309

RESUMO

Sedimentation velocity analytical ultracentrifugation (SV-AUC) is an indispensable tool for the study of particle size distributions in biopharmaceutical industry, for example, to characterize protein therapeutics and vaccine products. In particular, the diffusion-deconvoluted sedimentation coefficient distribution analysis, in the software SEDFIT, has found widespread applications due to its relatively high resolution and sensitivity. However, a lack of suitable software compatible with Good Manufacturing Practices (GMP) has hampered the use of SV-AUC in this regulatory environment. To address this, we have created an interface for SEDFIT so that it can serve as an automatically spawned module with controlled data input through command line parameters and output of key results in files. The interface can be integrated in custom GMP compatible software, and in scripts that provide documentation and meta-analyses for replicate or related samples, for example, to streamline analysis of large families of experimental data, such as binding isotherm analyses in the study of protein interactions. To test and demonstrate this approach we provide a MATLAB script mlSEDFIT.


Assuntos
Comércio , Documentação , Difusão , Registros , Software
20.
J Med Internet Res ; 25: e42047, 2023 09 06.
Artigo em Inglês | MEDLINE | ID: mdl-37672333

RESUMO

BACKGROUND: Predicting the likelihood of success of weight loss interventions using machine learning (ML) models may enhance intervention effectiveness by enabling timely and dynamic modification of intervention components for nonresponders to treatment. However, a lack of understanding and trust in these ML models impacts adoption among weight management experts. Recent advances in the field of explainable artificial intelligence enable the interpretation of ML models, yet it is unknown whether they enhance model understanding, trust, and adoption among weight management experts. OBJECTIVE: This study aimed to build and evaluate an ML model that can predict 6-month weight loss success (ie, ≥7% weight loss) from 5 engagement and diet-related features collected over the initial 2 weeks of an intervention, to assess whether providing ML-based explanations increases weight management experts' agreement with ML model predictions, and to inform factors that influence the understanding and trust of ML models to advance explainability in early prediction of weight loss among weight management experts. METHODS: We trained an ML model using the random forest (RF) algorithm and data from a 6-month weight loss intervention (N=419). We leveraged findings from existing explainability metrics to develop Prime Implicant Maintenance of Outcome (PRIMO), an interactive tool to understand predictions made by the RF model. We asked 14 weight management experts to predict hypothetical participants' weight loss success before and after using PRIMO. We compared PRIMO with 2 other explainability methods, one based on feature ranking and the other based on conditional probability. We used generalized linear mixed-effects models to evaluate participants' agreement with ML predictions and conducted likelihood ratio tests to examine the relationship between explainability methods and outcomes for nested models. We conducted guided interviews and thematic analysis to study the impact of our tool on experts' understanding and trust in the model. RESULTS: Our RF model had 81% accuracy in the early prediction of weight loss success. Weight management experts were significantly more likely to agree with the model when using PRIMO (χ2=7.9; P=.02) compared with the other 2 methods with odds ratios of 2.52 (95% CI 0.91-7.69) and 3.95 (95% CI 1.50-11.76). From our study, we inferred that our software not only influenced experts' understanding and trust but also impacted decision-making. Several themes were identified through interviews: preference for multiple explanation types, need to visualize uncertainty in explanations provided by PRIMO, and need for model performance metrics on similar participant test instances. CONCLUSIONS: Our results show the potential for weight management experts to agree with the ML-based early prediction of success in weight loss treatment programs, enabling timely and dynamic modification of intervention components to enhance intervention effectiveness. Our findings provide methods for advancing the understandability and trust of ML models among weight management experts.


Assuntos
Inteligência Artificial , Software , Humanos , Aprendizado de Máquina , Confiança , Redução de Peso
SELEÇÃO DE REFERÊNCIAS
DETALHE DA PESQUISA
...