Forecasting severe respiratory disease hospitalizations using machine learning algorithms.

BACKGROUND: Forecasting models predicting trends in hospitalization rates have the potential to inform hospital management during seasonal epidemics of respiratory diseases and the associated surges caused by acute hospital admissions. Hospital bed requirements for elective surgery could be better planned if it were possible to foresee upcoming peaks in severe respiratory illness admissions. Forecasting models can also guide the use of intervention strategies to decrease the spread of respiratory pathogens and thus prevent local health system overload. In this study, we explore the capability of forecasting models to predict the number of hospital admissions in Auckland, New Zealand, within a three-week time horizon. Furthermore, we evaluate probabilistic forecasts and the impact on model performance when integrating laboratory data describing the circulation of respiratory viruses. METHODS: The dataset used for this exploration results from active hospital surveillance, in which the World Health Organization Severe Acute Respiratory Infection (SARI) case definition was consistently used. This research nurse-led surveillance has been implemented in two public hospitals in Auckland and provides a systematic laboratory testing of SARI patients for nine respiratory viruses, including influenza, respiratory syncytial virus, and rhinovirus. The forecasting strategies used comprise automatic machine learning, one of the most recent generative pre-trained transformers, and established artificial neural network algorithms capable of univariate and multivariate forecasting. RESULTS: We found that machine learning models compute more accurate forecasts in comparison to naïve seasonal models. Furthermore, we analyzed the impact of reducing the temporal resolution of forecasts, which decreased the model error of point forecasts and made probabilistic forecasting more reliable. An additional analysis that used the laboratory data revealed strong season-to-season variations in the incidence of respiratory viruses and how this correlates with total hospitalization cases. These variations could explain why it was not possible to improve forecasts by integrating this data. CONCLUSIONS: Active SARI surveillance and consistent data collection over time enable these data to be used to predict hospital bed utilization. These findings show the potential of machine learning as support for informing systems for proactive hospital management.

Advancements in biotransformation pathway prediction: enhancements, datasets, and novel functionalities in enviPath.

enviPath is a widely used database and prediction system for microbial biotransformation pathways of primarily xenobiotic compounds. Data and prediction system are freely available both via a web interface and a public REST API. Since its initial release in 2016, we extended the data available in enviPath and improved the performance of the prediction system and usability of the overall system. We now provide three diverse data sets, covering microbial biotransformation in different environments and under different experimental conditions. This also enabled developing a pathway prediction model that is applicable to a more diverse set of chemicals. In the prediction engine, we implemented a new evaluation tailored towards pathway prediction, which returns a more honest and holistic view on the performance. We also implemented a novel applicability domain algorithm, which allows the user to estimate how well the model will perform on their data. Finally, we improved the implementation to speed up the overall system and provide new functionality via a plugin system. SCIENTIFIC CONTRIBUTION: The main scientific contributions are the development of a pathway prediction model applicable to diverse chemicals, a specialized evaluation method for holistic performance assessment, and a novel applicability domain algorithm for user-specific performance estimation. The introduction of two new data sets, and the creation of links to EC classes make enviPath a unique resource in microbial biotransformation research.

Immunity against measles, mumps, rubella, and varicella among homeless individuals in Germany - A nationwide multi-center cross-sectional study.

Introduction: Homeless individuals suffer a high burden of vaccine-preventable infectious diseases. Moreover, they are particularly susceptible to adverse infection outcomes with limited access to the health care system. Data on the seroprevalence of measles, mumps, rubella, and varicella within this cohort are missing. Methods: The seroprevalence of measles, mumps, rubella, and varicella was determined within the homeless population in Germany. Predictors of lacking immune protection were determined using multivariable logistic regression analysis. Results: Homeless individuals in Germany (n = 611) showed a seroprevalence of 88.5% (95% CI: 85.8-91.0) for measles, 83.8% (95% CI: 80.6-86.6) for mumps, 86.1% (95% CI: 83.1-88.7) for rubella, and 95.7% (95% CI 93.8-97.2) for varicella. Measles seroprevalences declined from individuals born in 1965 to individuals born in 1993, with seroprevalences not compatible with a 95% threshold in individuals born after 1980. For mumps, seroprevalences declined from individuals born in 1950 to individuals born in 1984. Here, seroprevalences were not compatible with a 92% threshold for individuals born after 1975. Seronegativity for measles, mumps and rubella was associated with age but not with gender or country of origin. Discussion: Herd immunity for measles and mumps is not achieved in this homeless cohort, while there was sufficient immune protection for rubella and varicella. Declining immune protection rates in younger individuals warrant immunization campaigns also targeting marginalized groups such as homeless individuals. Given that herd immunity thresholds are not reached for individuals born after 1980 for measles, and after 1975 for mumps, vaccination campaigns should prioritize individuals within these age groups.

AdductHunter: identifying protein-metal complex adducts in mass spectra.

Mass spectrometry (MS) is an analytical technique for molecule identification that can be used for investigating protein-metal complex interactions. Once the MS data is collected, the mass spectra are usually interpreted manually to identify the adducts formed as a result of the interactions between proteins and metal-based species. However, with increasing resolution, dataset size, and species complexity, the time required to identify adducts and the error-prone nature of manual assignment have become limiting factors in MS analysis. AdductHunter is a open-source web-based analysis tool that  automates the peak identification process using constraint integer optimization to find feasible combinations of protein and fragments, and dynamic time warping to calculate the dissimilarity between the theoretical isotope pattern of a species and its experimental isotope peak distribution. Empirical evaluation on a collection of 22 unique MS datasetsshows fast and accurate identification of protein-metal complex adducts in deconvoluted mass spectra.

Determinants of health-related quality of life (HRQoL) among homeless individuals during the COVID-19 pandemic.

OBJECTIVE: Thus far, there is very limited knowledge regarding homeless individuals during the COVID-19 pandemic, particularly related to the health-related quality of life (HRQoL). Thus, our aim was to evaluate HRQoL and to clarify the determinants of HRQoL among homeless individuals during the COVID-19 pandemic in Germany. METHODS: Data were taken from the national survey on psychiatric and somatic health of homeless individuals during the COVID-19 pandemic-NAPSHI (n = 616). The established EQ-5D-5L was used to quantify problems in five health dimensions, and its visual analogue scale (EQ-VAS) was used to record self-rated health status. Sociodemographic factors were included in regression analysis. RESULTS: Pain/discomfort was the most frequently reported problem (45.3%), thereafter anxiety/depression (35.9%), mobility (25.4%), usual activities (18.5%) and self-care (11.4%). Average EQ-VAS score was 68.97 (SD: 23.83), and the mean EQ-5D-5L index was 0.85 (SD: 0.24). Regressions showed that higher age and having a health insurance were associated with several problem dimensions. Being married was associated with higher EQ-VAS scores. CONCLUSIONS: Overall, our study findings showed a quite high HRQoL among homeless individuals during the COVID-19 pandemic in Germany. Some important determinants of HRQoL were identified (e.g., age or marital status). Longitudinal studies are required to confirm our findings.

Combatting over-specialization bias in growing chemical databases.

BACKGROUND: Predicting in advance the behavior of new chemical compounds can support the design process of new products by directing the research toward the most promising candidates and ruling out others. Such predictive models can be data-driven using Machine Learning or based on researchers' experience and depend on the collection of past results. In either case: models (or researchers) can only make reliable assumptions about compounds that are similar to what they have seen before. Therefore, consequent usage of these predictive models shapes the dataset and causes a continuous specialization shrinking the applicability domain of all trained models on this dataset in the future, and increasingly harming model-based exploration of the space. PROPOSED SOLUTION: In this paper, we propose CANCELS (CounterActiNg Compound spEciaLization biaS), a technique that helps to break the dataset specialization spiral. Aiming for a smooth distribution of the compounds in the dataset, we identify areas in the space that fall short and suggest additional experiments that help bridge the gap. Thereby, we generally improve the dataset quality in an entirely unsupervised manner and create awareness of potential flaws in the data. CANCELS does not aim to cover the entire compound space and hence retains a desirable degree of specialization to a specified research domain. RESULTS: An extensive set of experiments on the use-case of biodegradation pathway prediction not only reveals that the bias spiral can indeed be observed but also that CANCELS produces meaningful results. Additionally, we demonstrate that mitigating the observed bias is crucial as it cannot only intervene with the continuous specialization process, but also significantly improves a predictor's performance while reducing the number of required experiments. Overall, we believe that CANCELS can support researchers in their experimentation process to not only better understand their data and potential flaws, but also to grow the dataset in a sustainable way. All code is available under github.com/KatDost/Cancels .

The Mental and Physical Health of the Homeless.

BACKGROUND: The health status of homeless individuals in Germany has been described incompletely. Mental and somatic illnesses seem to contribute to the high mortality in this cohort. METHODS: In this national, multicenter, cross-sectional study, data were collected on the health of 651 homeless individuals in the metropolitan regions of Hamburg, Frankfurt, Leipzig, and Munich metropolitan regions. The lifetime prevalences of physician-diagnosed mental and somatic illnesses were determined with interviewbased questionnaires. Furthermore, clinical and laboratory examinations were carried out. Multivariable regressions were performed to identify determinants of health status and access to care. RESULTS: High prevalences of both mental and somatic illnesses were confirmed. Particularly, cardiovascular and metabolic diseases were highly prevalent. Evidence for possible unrecognized arterial hypertension and possible unrecognized hypercholesterolemia was found in 27.5% and 15.6% of homeless individuals, respec - tively. 23.1% of study participants reported having received a diagnosis of a mental illness. Evidence for a possible unrecognized mental illness was found in 69.7%. A history of immigration from another country to Germany was found to be an important determinant of the summed scores for mental, somatic, and possible unrecognized illness. Homeless individuals of non-German origin were more likely to be living without shelter (p = 0.03) and to lack health insurance (p < 0.001). CONCLUSION: High prevalence rates for mental and somatic illnesses and limited access to mainstream medical care were found. Homeless individuals appear to receive inadequate care for mental illness. Healthcare programs for homeless individuals in Germany should pay particular attention to homeless migrants.

Predictors of Loneliness among Homeless Individuals in Germany during the COVID-19 Pandemic.

PURPOSE: The aim of the study was to identify the frequency of loneliness and to examine the factors associated with loneliness among homeless individuals in Germany during the COVID-19 pandemic. METHODS: Data were taken from the 'national survey on the psychiatric and somatic health of homeless individuals during the COVID-19 pandemic'. The data collection took place from 26th July to 17th September 2021 (the analytical sample included n = 491 observations). The well-established UCLA-3 tool was used to quantify loneliness. Independent variables included sex, age, marital status, the existence of children and pets, level of education, country of origin, duration of homelessness, alcohol and drug consumption, mental health concerns and concerns regarding COVID-19 illness. Multiple logistic regressions were used to examine the predictors of loneliness. RESULTS: The frequency of loneliness was 41.7% for the total sample. Multiple logistic regression analysis stratified by gender showed that a higher likelihood of loneliness was associated with being born in Germany, being middle aged (40 to 49 years compared to 18 to 29 years), having mental health problems and a short period of homelessness (1 month compared to longer periods) among women. In men, a higher likelihood of loneliness was associated with a higher fear of COVID-19 and a short period of homelessness. CONCLUSIONS: Our study revealed a high frequency rate of loneliness among homeless individuals. The study results highlight the associations between some explanatory variables (i.e., the duration of homelessness and mental health problems). Identifying the factors associated with loneliness may help to adequately address the problems of homeless individuals at risk of loneliness. Longitudinal studies are required to confirm our findings.

Fear of COVID-19 among homeless individuals in Germany in mid-2021.

Aims: To investigate the prevalence and the correlates of fear of COVID-19 among homeless individuals. Methods: We used data from the "national survey on psychiatric and somatic health of homeless individuals during the COVID-19 pandemic" (NAPSHI-study) which took place in several large cities in Germany in Mid-2021 (n = 666 in the analytical sample). Mean age equaled 43.3 years (SD: 12.1 years), ranging from 18 to 80 years. Multiple linear regressions were performed. Results: In our study, 70.9% of the homeless individuals reported no fear of COVID-19. Furthermore, 14.0% reported a little fear of COVID-19, 8.4% reported some fear of COVID-19 and 6.7% reported severe fear of COVID-19. Multiple linear regressions revealed that fear of COVID-19 was higher among individuals aged 50-64 years (compared to individuals aged 18-29 years: ß = 0.28, p < 0.05), among individuals with a higher perceived own risk of contracting the coronavirus 1 day (ß = 0.28, p < 0.001) as well as among individuals with a higher agreement that a diagnosis of the coronavirus would ruin his/her life (ß = 0.15, p < 0.001). Conclusions: Only a small proportion of homeless individuals reported fear of COVID-19 in mid-2021 in Germany. Such knowledge about the correlates of higher levels of fear of COVID-19 may be helpful for addressing certain risk groups (e.g., homeless individuals aged 50-64 years). In a further step, avoiding extraordinarily high levels of fear of COVID-19 may be beneficial to avoid irrational thinking and acting regarding COVID-19 in this group.