Florida Museum of Natural History

BackgroundOpioid use disorder (OUD) remains a critical public health crisis in the United States. Despite widespread policy and clinical interventions, early identification of individuals at risk for developing OUD remains challenging due to limitations in traditional screening approaches and a lack of individualized risk stratification methods. Machine learning (ML) methods offer an opportunity to develop timely, high-performing, and explainable predictive models that can enhance OUD prevention strategies in clinical settings.ObjectiveThis study aims to develop and validate an ML model using electronic health record (EHR) data to predict the 3-month risk of incident OUD among adults initiating opioid therapy and to stratify patients into clinically actionable risk groups.MethodsThis prognostic modeling study used 2017‐2022 OneFlorida+ EHR data to develop and validate ML algorithms predicting 3-month incident OUD risk. We included 182,083 adults (≥18 y) without cancer, overdose, or OUD or hospice history who received ≥1 outpatient, noninjectable opioid prescription. Using 183 predictors measured in sequential 3-month intervals, we developed an elastic net, least absolute shrinkage and selection operator, gradient boosting machine (GBM), and random forest models on randomly split training, testing, and validation sets. Model performance was assessed using C-statistics, predictive values, and number needed to evaluate, with patients stratified into risk deciles for clinical applicability. Model explainability was assessed using Shapley additive explanations, and fairness was evaluated using standard metrics. We externally validated the best-performing model using an independent cohort from the 2018‐2020 UPMC (formerly University of Pittsburgh Medical Center) health system.ResultsIn the validation sample (n=60,694), GBM (C-statistics=0.879, 95% CI 0.874‐0.884) and elastic net (C-statistics=0.872, 95% CI 0.867‐0.877) outperformed least absolute shrinkage and selection operator (C-statistics=0.846, 95% CI 0.840‐0.851) and random forest (C-statistics=0.798, 95% CI 0.792‐0.804), with GBM model requiring the fewest predictors (n=75) for predicting 3-month incident OUD. Using the GBM algorithm to predict the subsequent 3-month OUD risk, the top decile subgroup had a positive predictive value of 3.26%, a negative predictive value of 99.8%, and a number needed to evaluate of 31. The top decile (n=6696) captured ~68% of patients with OUD. Shapley additive explanations analysis identified age, number of outpatient visits, history of back and other pain conditions, comorbidity burden, and opioid prescribing patterns as the strongest predictors of incident OUD. Fairness assessment showed an acceptable false negative rate parity across race, age, and sex. In external validation on the UPMC cohort, the GBM model maintained good discrimination (C-statistics=0.756, 95% CI 0.750‐0.762) and effective risk stratification.ConclusionsAn ML algorithm predicting incident OUD derived from OneFlorida+ EHR data performed well in external validation with data using UPMC. The algorithm might be valuable for incident OUD risk prediction and stratification across health systems, with potential to inform early intervention.

BackgroundOpioids are a widely prescribed class of medication for pain management. However, they have variable efficacy and adverse effects among patients, due to the complex interplay between biological and clinical factors. Pharmacogenetic testing can be used to match patients’ genetic profiles to individualize opioid therapy, improving pain relief and reducing the risk of adverse effects. Despite its potential, the pharmacogenetic testing uptake (use of pharmacogenetic testing) remains low due to a range of barriers at the patient, health care provider, infrastructure, and financial levels. Since testing typically involves a shared decision between the provider and patient, predicting the likelihood of a patient undergoing pharmacogenetic testing and understanding the factors influencing that decision can help optimize resource use and improve outcomes in pain management.ObjectiveThis study aimed to develop machine learning (ML) models, identifying patients’ likelihood of pharmacogenetic uptake based on their demographics, clinical variables, medication use, and social determinants of health.MethodsWe used electronic health record data from a single center health care system to identify patients prescribed opioids. We extracted patients’ demographics, clinical variables, medication use, and social determinants of health, and developed and validated ML models, including a neural network, logistic regression, random forest, extreme gradient boosting (XGB), naïve Bayes, and support vector machines for pharmacogenetic testing uptake prediction based on procedure codes. We performed 5-fold cross-validation and created an ensemble probability-based classifier using the best-performing ML models for pharmacogenetic testing uptake prediction. Various performance metrics, uptake stratification analysis, and feature importance analysis were used to evaluate the performance of the models.ResultsThe ensemble model using XGB and support vector machine–radial basis function classifiers had the highest C-statistics at 79.61%, followed by XGB (78.94%), and neural network (78.05%). While XGB was the best-performing model, the ensemble model achieved a high accuracy (32,699/48,528, 67.38%), recall (537/702, 76.50%), specificity (32,162/47,826, 67.25%), and negative predictive value (32,162/32,327, 99.49%). The uptake stratification analysis using the ensemble model indicated that it can effectively distinguish across uptake probability deciles, where those in the higher strata are more likely to undergo pharmacogenetic testing in the real world (320/4853, 6.59% in the highest decile compared to 6/4853, 0.12% in the lowest). Furthermore, Shapley Additive Explanations value analysis using the XGB model indicated age, hypertension, and household income as the most influential factors for pharmacogenetic testing uptake prediction.ConclusionsThe proposed ensemble model demonstrated a high performance in pharmacogenetic testing uptake prediction among patients using opioids for pain. This model can be used as a decision support tool, assisting clinicians in identifying patients’ likelihood of pharmacogenetic testing uptake and guiding appropriate decision-making.

Premise: Molecular datasets for estimating phylogenetic trees increasingly include more species and gene regions. Often trees are constructed using backbone phylogenies, subtrees, and other techniques to address the challenges of large dataset size. Currently, there is no established approach to integrate these rapidly expanding datasets. Methods: We generated a phylogenetic tree (and 1,000 bootstrap trees) with divergence times that span euphyllophytes. To do this, we integrated taxonomically broad dated backbone phylogenies with species-level trees generated from phylogenetic analysis of individual clades. Datasets for species-level trees were assembled using PyPHLAWD. Results: The resulting dated phylogenetic tree includes: 121,641 angiosperm species; 1,026 gymnosperms; and 5,603 ferns. This is the largest euphyllophyte phylogenetic tree constructed to date. Topological uncertainty spikes at the start of the Cretaceous and gradually increases during the Cenozoic. Uncertainty in age estimates gradually increases in the Cenozoic but increases dramatically in the most recent 5 Myrs. Discussion: This dated phylogenetic framework can underpin evolutionary studies spanning euphyllophytes, and enable the integration of insights from the recent to the distant past. Our approach also enables the tree to be easily updated in the future to reflect future increases in data availability, and systematic and taxonomic advances within specific clades.

ABSTRACT Aim Several mechanisms contribute to the plant biodiversity of the Neotropics, with the highlands of South America serving as important hotspots of diversity. In particular, the Brazilian highlands exhibit high biodiversity due to complex diversification dynamics and a mixture of contributions from different biomes. In this study, we reconstruct the timing and potential triggers of diversification of Agalinis , hemiparasitic plants that inhabit open grassland habitats, to investigate their biogeographic history and migration patterns across the Americas. Location North, Central and South America. Taxon Agalinis . Methods We reconstructed dated phylogenies of Agalinis using a secondary calibration approach, sampling 73% of the known species, including multiple species from the Andes and Brazilian highlands. We inferred ancestral distributions to understand migration patterns between North and South America and within South America. Additionally, we investigated shifts in diversification rates within the genus and reconstructed ancestral pollination syndrome and life‐history strategy states. All analyses were performed across a distribution of trees to account for phylogenetic uncertainty. Results Agalinis likely originated in south‐eastern North America during the early Miocene and rapidly diversified, followed by movement into South America in the Late Miocene or Early Pliocene. We propose two possible routes for Agalinis movement into South America: either through the Andes or via the South American lowland grasslands (e.g., Chaco, Pampas, Cerrado, Caatinga, Llanos), using grassland corridors for dispersal within the continent. After its arrival in South American highlands, the clade underwent rapid diversification. State reconstructions indicated that the genus had a bee‐pollinated ancestor and that hummingbird pollination evolved only once, with many transitions back to bee pollination. In contrast, the perennial life strategy evolved multiple times within the genus, including at least once in the ancestor of all South American species and twice in the North American species. Main Conclusions Agalinis likely originated in North America and later migrated to South America, followed by rapid diversification (i.e., ‘dispersification’). Thus, Agalinis is a clade that refutes the tropical conservatism hypothesis and the out‐of‐the‐tropics model. Moreover, the dispersal to the Brazilian highlands was from nearby species pools and not from other highland habitats such as the Andes, followed by several in situ speciation events. These high diversification rates were partly associated with Quaternary climatic oscillations, and perennial and hummingbird‐pollinated species exhibited higher turnover rates.

Abstract Predation is a key driver of speciation and phenotypic diversification, yet how antipredator traits evolve and persist over evolutionary time remains poorly understood. We generated target sequence capture and whole-genome sequencing data for the skipper butterfly subfamily Eudaminae (Hesperiidae) to test whether antipredator defenses, in interaction with environmental variation, promote diversification. We focus on two antipredation wing traits: hindwing tails, which deflect predator attacks, and blue-green coloration, which may enhance motion dazzle and be used as warning coloration. Applying phylogenomics, morphometrics, and comparative methods, we model trait evolution in relation to diel activity and geographic distribution and find that hindwing tails repeatedly evolved at least seven times and blue-green coloration at least fifteen times. Both traits are associated with elevated speciation rates, but evolutionary transitions toward tailless wings and non-iridescent coloration occurred more frequently than trait gains, indicating high evolutionary lability of these antipredator defenses. Trait loss, particularly pronounced in tropical diurnal species, may reflect trade-offs in flight performance, shifts in predation guilds, or the evolution of alternative defensive traits. Our findings highlight rampant convergent evolution of wing traits that are under strong predator-mediated selection. By identifying how antipredator defenses and environmental contexts influence phenotypic and species diversification, this study provides new insights into the ecological and evolutionary processes underlying insect diversity.

This leading textbook introduces students and practitioners to the identification and analysis of animal remains at archaeology sites. The authors use global examples from the Pleistocene era into the present to explain how zooarchaeology allows us to form insights about relationships among people and their natural and social environments, especially site-formation processes, economic strategies, domestication, and paleoenvironments. This new edition reflects the significant technological developments in zooarchaeology that have occurred in the past two decades, notably ancient DNA, proteomics, and isotope geochemistry. Substantially revised to reflect these trends, the volume also highlights novel applications, current issues in the field, the growth of international zooarchaeology, and the increased role of interdisciplinary collaborations. In view of the growing importance of legacy collections, voucher specimens, and access to research materials, it also includes a substantially revised chapter that addresses management of zooarchaeological collections and curation of data.

Radiogenic strontium (87Sr/86Sr) isotope analysis is a well-established method for reconstructing the mobility of human populations in the past and present. Baseline 87Sr/86Sr data are fundamental to the method, as Sr varies across the landscape according to local geology and geoenvironmental factors. The method's application within studies of ancient Mesoamerican paleomobility, however, has concentrated on two key regions-Teotihuacan and the Maya region-despite its potential broader relevance across greater Mesoamerica. This is due in part to a lack of available baseline 87Sr/86Sr data for the region at large. Using the Valley of Oaxaca as a case study, we use Bayesian Additive Regression Trees (BART) to generate a locally calibrated predictive 87Sr/86Sr isoscape model of Mesoamerica in general and the Valley of Oaxaca in particular. We integrate (1) observed 87Sr/86Sr data from modern plant samples (n = 95) from 17 sites across the Valley, (2) a compiled database of continental North and South American 87Sr/86Sr data, (3) geological bedrock maps, and (4) high resolution spatial data on geoenvironmental Sr covariates to iteratively develop and test a high performing predictive model for Mesoamerica, highlighting the importance of regional calibration in developing predictive 87Sr/86Sr isoscapes. Our results indicate that though overlap exists, 87Sr/86Sr can be used to detect migration within the Valley of Oaxaca as well as between the Valley and greater Mesoamerica. We then apply our isoscape to previously published human 87Sr/86Sr data from Monte Albán, Oaxaca to demonstrate how our model's explicit quantification of uncertainty in local 87Sr/86Sr ranges allows for more nuanced interpretation of paleomobility in archaeological samples.

Two new feather mite species of the genus Proterothrix Gaud, 1968 (Proctophyllodidae: Pterodectinae) collected in China from the Red-tailed Laughingthrush Trochalopteron milnei (Passeriformes: Leiothrichidae) are described: Proterothrix papilio sp. n. and P. nanduhensis sp. n. Both new species belong to the paradoxornis species group in having in males seta e of tarsus I lanceolate and aedeagus long whip-shaped. Males of P. papilio have a unique distinctive character within the genus: a pair of stick-like pregenital sclerites; in other species of the genus this sclerite, if present, is represented by single median sclerite. Females of this species have setae h2 with a terminal filament, setae h1 situated at the level of supranal concavity, and genual setae mGI, mGII thick spiculiform. Males of P. nanduhensis have the aedeagus extending beyond lobar apices by almost half its length and the opisthosomal lobes elongated, attenuate to apex, with posterior ends rounded. Females of this species have lateral margins of prodorsal shield with small round incisions in anterior quarter, at level of trochanters I.