Foresight-a generative pretrained transformer for modelling of patient timelines using electronic health records: a retrospective modelling study.

BACKGROUND: An electronic health record (EHR) holds detailed longitudinal information about a patient's health status and general clinical history, a large portion of which is stored as unstructured, free text. Existing approaches to model a patient's trajectory focus mostly on structured data and a subset of single-domain outcomes. This study aims to evaluate the effectiveness of Foresight, a generative transformer in temporal modelling of patient data, integrating both free text and structured formats, to predict a diverse array of future medical outcomes, such as disorders, substances (eg, to do with medicines, allergies, or poisonings), procedures, and findings (eg, relating to observations, judgements, or assessments). METHODS: Foresight is a novel transformer-based pipeline that uses named entity recognition and linking tools to convert EHR document text into structured, coded concepts, followed by providing probabilistic forecasts for future medical events, such as disorders, substances, procedures, and findings. The Foresight pipeline has four main components: (1) CogStack (data retrieval and preprocessing); (2) the Medical Concept Annotation Toolkit (structuring of the free-text information from EHRs); (3) Foresight Core (deep-learning model for biomedical concept modelling); and (4) the Foresight web application. We processed the entire free-text portion from three different hospital datasets (King's College Hospital [KCH], South London and Maudsley [SLaM], and the US Medical Information Mart for Intensive Care III [MIMIC-III]), resulting in information from 811 336 patients and covering both physical and mental health institutions. We measured the performance of models using custom metrics derived from precision and recall. FINDINGS: Foresight achieved a precision@10 (ie, of 10 forecasted candidates, at least one is correct) of 0·68 (SD 0·0027) for the KCH dataset, 0·76 (0·0032) for the SLaM dataset, and 0·88 (0·0018) for the MIMIC-III dataset, for forecasting the next new disorder in a patient timeline. Foresight also achieved a precision@10 value of 0·80 (0·0013) for the KCH dataset, 0·81 (0·0026) for the SLaM dataset, and 0·91 (0·0011) for the MIMIC-III dataset, for forecasting the next new biomedical concept. In addition, Foresight was validated on 34 synthetic patient timelines by five clinicians and achieved a relevancy of 33 (97% [95% CI 91-100]) of 34 for the top forecasted candidate disorder. As a generative model, Foresight can forecast follow-on biomedical concepts for as many steps as required. INTERPRETATION: Foresight is a general-purpose model for biomedical concept modelling that can be used for real-world risk forecasting, virtual trials, and clinical research to study the progression of disorders, to simulate interventions and counterfactuals, and for educational purposes. FUNDING: National Health Service Artificial Intelligence Laboratory, National Institute for Health and Care Research Biomedical Research Centre, and Health Data Research UK.

Evolution of reproductive life-history and dispersal traits during the range expansion of a biological control agent.

Evolutionary theory predicts that the process of range expansion will lead to differences in life-history and dispersal traits between the core and edge of a population. At the edge, selection and genetic drift can have opposing effects on reproductive ability, while spatial sorting by dispersal ability can increase dispersal. However, the context that individuals experience, including population density and mating status, also impacts dispersal behavior. We seek to understand the shifts in traits of populations expanding across natural, heterogenous environments, and the evolutionary and behavioral factors that may drive those shifts. We evaluated theoretical predictions for evolution of reproductive life-history and dispersal traits using the range expansion of a biological control agent, Diorhabda carinulata, or northern tamarisk beetle. We find that individuals from the edge had increased fecundity and female body mass, and reduced age at first reproduction, indicating that genetic load is low and suggesting that selection has acted at the edge. We also find that density of conspecifics during rearing and mating status influence dispersal of males and that dispersal increases at the edge of the range under certain conditions, particularly when males were unmated and reared at low density. The restricted conditions in which dispersal has increased suggest that spatial sorting has exerted weak effects relative to other potential processes. Our results support most theoretical predictions about evolution during range expansion, even across a heterogeneous environment, especially when the ecological context is considered.

Hybridization and range expansion in tamarisk beetles (Diorhabda spp.) introduced to North America for classical biological control.

With the global rise of human-mediated translocations and invasions, it is critical to understand the genomic consequences of hybridization and mechanisms of range expansion. Conventional wisdom is that high genetic drift and loss of genetic diversity due to repeated founder effects will constrain introduced species. However, reduced genetic variation can be countered by behavioral aspects and admixture with other distinct populations. As planned invasions, classical biological control (biocontrol) agents present important opportunities to understand the mechanisms of establishment and spread in a novel environment. The ability of biocontrol agents to spread and adapt, and their effects on local ecosystems, depends on genomic variation and the consequences of admixture in novel environments. Here, we use a biocontrol system to examine the genome-wide outcomes of introduction, spread, and hybridization in four cryptic species of a biocontrol agent, the tamarisk beetle (Diorhabda carinata, D. carinulata, D. elongata, and D. sublineata), introduced from six localities across Eurasia to control the invasive shrub tamarisk (Tamarix spp.) in western North America. We assembled a de novo draft reference genome and applied RADseq to over 500 individuals across laboratory cultures, the native ranges, and the introduced range. Despite evidence of a substantial genetic bottleneck among D. carinulata in N. America, populations continue to establish and spread, possibly due to aggregation behavior. We found that D. carinata, D. elongata, and D. sublineata hybridize in the field to varying extents, with D. carinata × D. sublineata hybrids being the most abundant. Genetic diversity was greater at sites with hybrids, highlighting potential for increased ability to adapt and expand. Our results demonstrate the complex patterns of genomic variation that can result from introduction of multiple ecotypes or species for biocontrol, and the importance of understanding them to predict and manage the effects of biocontrol agents in novel ecosystems.

Using Chemical Ecology to Enhance Weed Biological Control.

In agricultural systems, chemical ecology and the use of semiochemicals have become critical components of integrated pest management. The categories of semiochemicals that have been used include sex pheromones, aggregation pheromones, and plant volatile compounds used as attractants as well as repellents. In contrast, semiochemicals are rarely utilized for management of insects used in weed biological control. Here, we advocate for the benefit of chemical ecology principles in the implementation of weed biocontrol by describing successful utilization of semiochemicals for release, monitoring and manipulation of weed biocontrol agent populations. The potential for more widespread adoption and successful implementation of semiochemicals justifies multidisciplinary collaborations and increased research on how semiochemicals and chemical ecology can enhance weed biocontrol programs.

The Effects of ARBs, ACEis, and Statins on Clinical Outcomes of COVID-19 Infection Among Nursing Home Residents.

OBJECTIVES: Angiotensin-converting enzyme inhibitors (ACEi), angiotensin II receptor blockers (ARBs), and HMG-CoA reductase inhibitors ("statins") have been hypothesized to affect COVID-19 severity. However, up to now, no studies investigating this association have been conducted in the most vulnerable and affected population groups (ie, older adults residing in nursing homes). The objective of this study was to explore the association of ACEi/ARB and/or statins with clinical manifestations in COVID-19-infected older adults residing in nursing homes. DESIGN: We undertook a retrospective multicenter cohort study to analyze the association between ACEi/ARB and/or statin use with clinical outcome of COVID-19. The outcomes were (1) serious COVID-19 defined as long-stay hospital admission or death within 14 days of disease onset, and (2) asymptomatic (ie, no disease symptoms in the whole study period while still being diagnosed by polymerase chain reaction). SETTING AND PARTICIPANTS: A total of 154 COVID-19-positive subjects were identified, residing in 1 of 2 Belgian nursing homes that experienced similar COVID-19 outbreaks. MEASURES: Logistic regression models were applied with age, sex, functional status, diabetes, and hypertension as covariates. RESULTS: We found a statistically significant association between statin intake and the absence of symptoms during COVID-19 (odds ratio [OR] 2.91; confidence interval [CI] 1.27-6.71), which remained statistically significant after adjusting for covariates (OR 2.65; CI 1.13-6.68). Although the effects of statin intake on serious clinical outcome were in the same beneficial direction, these were not statistically significant (OR 0.75; CI 0.24-1.87). There was also no statistically significant association between ACEi/ARB and asymptomatic status (OR 2.72; CI 0.59-25.1) or serious clinical outcome (OR 0.48; CI 0.10-1.97). CONCLUSIONS AND IMPLICATIONS: Our data indicate that statin intake in older, frail adults could be associated with a considerable beneficial effect on COVID-19 clinical symptoms. The role of statins and renin-angiotensin system drugs needs to be further explored in larger observational studies as well as randomized clinical trials.

Network Approaches to the Understanding of Alzheimer's Disease: From Model Organisms to Humans.

It is becoming increasingly evident that Alzheimer's disease cannot be considered as the outcome of a single pathway, but rather we should view it as a system, that is, a network of interactions between large numbers of different protein molecules. In the last few years, probably because of the inherent limitations of traditional methods and because of the great increase in availability of sequencing data, this type of approach is being used more and more. In the following, we will discuss what constitutes a "network approach," what are its pros and cons, a number of recent case studies and finally what are the future perspectives of this type of analysis.

Species Introductions and Their Cascading Impacts on Biotic Interactions in desert riparian ecosystems.

Desert riparian ecosystems of North America are hotspots of biodiversity that support many sensitive species, and are in a region experiencing some of the highest rates of climatic alteration in North America. Fremont cottonwood, Populus fremontii, is a foundation tree species of this critical habitat, but it is threatened by global warming and regional drying, and by a non-native tree/shrub, Tamarix spp., all of which can disrupt the mutualism between P. fremontii and its beneficial mycorrhizal fungal communities. Specialist herbivorous leaf beetles (Diorhabda spp.) introduced for biocontrol of Tamarix are altering the relationship between this shrub and its environment. Repeated episodic feeding on Tamarix foliage by Diorhabda results in varying rates of dieback and mortality, depending on genetic variation in allocation of resources, growing conditions, and phenological synchrony between herbivore and host plant. In this article, we review the complex interaction between climatic change and species introductions and their combined impacts on P. fremontii and their associated communities. We anticipate that (1) certain genotypes of P. fremontii will respond more favorably to the presence of Tamarix and to climatic change due to varying selection pressures to cope with competition and stress; (2) the ongoing evolution of Diorhabda's life cycle timing will continue to facilitate its expansion in North America, and will over time enhance herbivore impact to Tamarix; (3) defoliation by Diorhabda will reduce the negative impact of Tamarix on P. fremontii associations with mycorrhizal fungi; and (4) spatial variability in climate and climatic change will modify the capacity for Tamarix to survive episodic defoliation by Diorhabda, thereby altering the relationship between Tamarix and P. fremontii, and its associated mycorrhizal fungal communities. Given the complex biotic/abiotic interactions outlined in this review, conservation biologists and riparian ecosystem managers should strive to identify and conserve the phenotypic traits that underpin tolerance and resistance to stressors such as climate change and species invasion. Such efforts will greatly enhance conservation restoration efficacy for protecting P. fremontii forests and their associated communities.

Life history characteristics of diorhabda carinulata under various temperatures.

Tamarisk leaf beetles, Diorhabda spp., have been released in the western United States as a biological control agent for the invasive weed Tamarix spp. There have been a few studies on the life cycle, host preferences, and field observations of Diorhabda; however, their ecophysiological characteristics under various temperature regimes are not clearly understood. In this study, life history characteristics such as growth, fecundity, and mortality of Diorhabda Carinulata (Desbrochers), the species established in the Colorado River basin, were investigated under various temperatures. Beetles were housed at various temperatures (room, constant high, and variable high) and their life cycle from eggs to reproductive adult was observed. Body size at various larval and adult stages, as well as their developmental time decreased with increasing temperature. Between the two temperature treatments, beetles at diurnally fluctuating temperature (variable high treatment) grew slower and produced fewer eggs per clutch when compared with the constant high treatment. Despite smaller in size, beetles grew fastest at the constant high temperature and produced most eggs per clutch compared with the other two treatments. Overall, severely high temperatures seem to have a debilitating effect on Diorhabda at early larval stages with nearly 50% mortality. The study has potential implications for the tamarisk beetle biocontrol program in the southwestern United States.

Evolution of critical day length for diapause induction enables range expansion of Diorhabda carinulata, a biological control agent against tamarisk (Tamarix spp.).

In classical weed biological control, small collections of arthropods are made from one or a few sites in the native range of the target plant and are introduced to suppress the plant where it has become invasive, often across a wide geographic range. Ecological mismatches in the new range are likely, and success using the biocontrol agent may depend on postrelease evolution of beneficial life history traits. In this study, we measure the evolution of critical day length for diapause induction (day length at which 50% of the population enters dormancy), in a beetle (Diorhabda carinulata) introduced into North America from China to control an exotic shrub, Tamarix spp. Beetle populations were sampled from four sites in North America 7 years after introduction, and critical day length was shown to have declined, forming a cline over a latitudinal gradient At one field site, decreased critical day length was correlated with 16 additional days of reproductive activity, resulting in a closer match between beetle life history and the phenology of Tamarix. These findings indicate an enhanced efficacy and an increasingly wider range for D. carinulata in Tamarix control.