Synthetic data generation for a longitudinal cohort study - evaluation, method extension and reproduction of published data analysis results.

Access to individual-level health data is essential for gaining new insights and advancing science. In particular, modern methods based on artificial intelligence rely on the availability of and access to large datasets. In the health sector, access to individual-level data is often challenging due to privacy concerns. A promising alternative is the generation of fully synthetic data, i.e., data generated through a randomised process that have similar statistical properties as the original data, but do not have a one-to-one correspondence with the original individual-level records. In this study, we use a state-of-the-art synthetic data generation method and perform in-depth quality analyses of the generated data for a specific use case in the field of nutrition. We demonstrate the need for careful analyses of synthetic data that go beyond descriptive statistics and provide valuable insights into how to realise the full potential of synthetic datasets. By extending the methods, but also by thoroughly analysing the effects of sampling from a trained model, we are able to largely reproduce significant real-world analysis results in the chosen use case.

Unraveling progression subtypes in people with Huntington's disease.

Background: Huntington's disease (HD) is a progressive neurodegenerative disease caused by a CAG trinucleotide expansion in the huntingtin gene. The length of the CAG repeat is inversely correlated with disease onset. HD is characterized by hyperkinetic movement disorder, psychiatric symptoms, and cognitive deficits, which greatly impact patient's quality of life. Despite this clear genetic course, high variability of HD patients' symptoms can be observed. Current clinical diagnosis of HD solely relies on the presence of motor signs, disregarding the other important aspects of the disease. By incorporating a broader approach that encompasses motor as well as non-motor aspects of HD, predictive, preventive, and personalized (3P) medicine can enhance diagnostic accuracy and improve patient care. Methods: Multisymptom disease trajectories of HD patients collected from the Enroll-HD study were first aligned on a common disease timescale to account for heterogeneity in disease symptom onset and diagnosis. Following this, the aligned disease trajectories were clustered using the previously published Variational Deep Embedding with Recurrence (VaDER) algorithm and resulting progression subtypes were clinically characterized. Lastly, an AI/ML model was learned to predict the progression subtype from only first visit data or with data from additional follow-up visits. Results: Results demonstrate two distinct subtypes, one large cluster (n = 7122) showing a relative stable disease progression and a second, smaller cluster (n = 411) showing a dramatically more progressive disease trajectory. Clinical characterization of the two subtypes correlates with CAG repeat length, as well as several neurobehavioral, psychiatric, and cognitive scores. In fact, cognitive impairment was found to be the major difference between the two subtypes. Additionally, a prognostic model shows the ability to predict HD subtypes from patients' first visit only. Conclusion: In summary, this study aims towards the paradigm shift from reactive to preventive and personalized medicine by showing that non-motor symptoms are of vital importance for predicting and categorizing each patients' disease progression pattern, as cognitive decline is oftentimes more reflective of HD progression than its motor aspects. Considering these aspects while counseling and therapy definition will personalize each individuals' treatment. The ability to provide patients with an objective assessment of their disease progression and thus a perspective for their life with HD is the key to improving their quality of life. By conducting additional analysis on biological data from both subtypes, it is possible to gain a deeper understanding of these subtypes and uncover the underlying biological factors of the disease. This greatly aligns with the goal of shifting towards 3P medicine. Supplementary Information: The online version contains supplementary material available at 10.1007/s13167-024-00368-2.

Towards early diagnosis of Alzheimer's disease: advances in immune-related blood biomarkers and computational approaches.

Alzheimer's disease has an increasing prevalence in the population world-wide, yet current diagnostic methods based on recommended biomarkers are only available in specialized clinics. Due to these circumstances, Alzheimer's disease is usually diagnosed late, which contrasts with the currently available treatment options that are only effective for patients at an early stage. Blood-based biomarkers could fill in the gap of easily accessible and low-cost methods for early diagnosis of the disease. In particular, immune-based blood-biomarkers might be a promising option, given the recently discovered cross-talk of immune cells of the central nervous system with those in the peripheral immune system. Here, we give a background on recent advances in research on brain-immune system cross-talk in Alzheimer's disease and review machine learning approaches, which can combine multiple biomarkers with further information (e.g. age, sex, APOE genotype) into predictive models supporting an earlier diagnosis. In addition, mechanistic modeling approaches, such as agent-based modeling open the possibility to model and analyze cell dynamics over time. This review aims to provide an overview of the current state of immune-system related blood-based biomarkers and their potential for the early diagnosis of Alzheimer's disease.

Progression subtypes in Parkinson's disease identified by a data-driven multi cohort analysis.

The progression of Parkinson's disease (PD) is heterogeneous across patients, affecting counseling and inflating the number of patients needed to test potential neuroprotective treatments. Moreover, disease subtypes might require different therapies. This work uses a data-driven approach to investigate how observed heterogeneity in PD can be explained by the existence of distinct PD progression subtypes. To derive stable PD progression subtypes in an unbiased manner, we analyzed multimodal longitudinal data from three large PD cohorts and performed extensive cross-cohort validation. A latent time joint mixed-effects model (LTJMM) was used to align patients on a common disease timescale. Progression subtypes were identified by variational deep embedding with recurrence (VaDER). In each cohort, we identified a fast-progressing and a slow-progressing subtype, reflected by different patterns of motor and non-motor symptoms progression, survival rates, treatment response, features extracted from DaTSCAN imaging and digital gait assessments, education, and Alzheimer's disease pathology. Progression subtypes could be predicted with ROC-AUC up to 0.79 for individual patients when a one-year observation period was used for model training. Simulations demonstrated that enriching clinical trials with fast-progressing patients based on these predictions can reduce the required cohort size by 43%. Our results show that heterogeneity in PD can be explained by two distinct subtypes of PD progression that are stable across cohorts. These subtypes align with the brain-first vs. body-first concept, which potentially provides a biological explanation for subtype differences. Our predictive models will enable clinical trials with significantly lower sample sizes by enriching fast-progressing patients.

The paradigm change from reactive medical services to 3PM in ischemic stroke: a holistic approach utilising tear fluid multi-omics, mitochondria as a vital biosensor and AI-based multi-professional data interpretation.

Worldwide stroke is the second leading cause of death and the third leading cause of death and disability combined. The estimated global economic burden by stroke is over US$891 billion per year. Within three decades (1990-2019), the incidence increased by 70%, deaths by 43%, prevalence by 102%, and DALYs by 143%. Of over 100 million people affected by stroke, about 76% are ischemic stroke (IS) patients recorded worldwide. Contextually, ischemic stroke moves into particular focus of multi-professional groups including researchers, healthcare industry, economists, and policy-makers. Risk factors of ischemic stroke demonstrate sufficient space for cost-effective prevention interventions in primary (suboptimal health) and secondary (clinically manifested collateral disorders contributing to stroke risks) care. These risks are interrelated. For example, sedentary lifestyle and toxic environment both cause mitochondrial stress, systemic low-grade inflammation and accelerated ageing; inflammageing is a low-grade inflammation associated with accelerated ageing and poor stroke outcomes. Stress overload, decreased mitochondrial bioenergetics and hypomagnesaemia are associated with systemic vasospasm and ischemic lesions in heart and brain of all age groups including teenagers. Imbalanced dietary patterns poor in folate but rich in red and processed meat, refined grains, and sugary beverages are associated with hyperhomocysteinaemia, systemic inflammation, small vessel disease, and increased IS risks. Ongoing 3PM research towards vulnerable groups in the population promoted by the European Association for Predictive, Preventive and Personalised Medicine (EPMA) demonstrates promising results for the holistic patient-friendly non-invasive approach utilising tear fluid-based health risk assessment, mitochondria as a vital biosensor and AI-based multi-professional data interpretation as reported here by the EPMA expert group. Collected data demonstrate that IS-relevant risks and corresponding molecular pathways are interrelated. For examples, there is an evident overlap between molecular patterns involved in IS and diabetic retinopathy as an early indicator of IS risk in diabetic patients. Just to exemplify some of them such as the 5-aminolevulinic acid/pathway, which are also characteristic for an altered mitophagy patterns, insomnia, stress regulation and modulation of microbiota-gut-brain crosstalk. Further, ceramides are considered mediators of oxidative stress and inflammation in cardiometabolic disease, negatively affecting mitochondrial respiratory chain function and fission/fusion activity, altered sleep-wake behaviour, vascular stiffness and remodelling. Xanthine/pathway regulation is involved in mitochondrial homeostasis and stress-driven anxiety-like behaviour as well as molecular mechanisms of arterial stiffness. In order to assess individual health risks, an application of machine learning (AI tool) is essential for an accurate data interpretation performed by the multiparametric analysis. Aspects presented in the paper include the needs of young populations and elderly, personalised risk assessment in primary and secondary care, cost-efficacy, application of innovative technologies and screening programmes, advanced education measures for professionals and general population-all are essential pillars for the paradigm change from reactive medical services to 3PM in the overall IS management promoted by the EPMA.

Redefining pandemic preparedness: Multidisciplinary insights from the CERP modelling workshop in infectious diseases, workshop report.

In July 2023, the Center of Excellence in Respiratory Pathogens organized a two-day workshop on infectious diseases modelling and the lessons learnt from the Covid-19 pandemic. This report summarizes the rich discussions that occurred during the workshop. The workshop participants discussed multisource data integration and highlighted the benefits of combining traditional surveillance with more novel data sources like mobility data, social media, and wastewater monitoring. Significant advancements were noted in the development of predictive models, with examples from various countries showcasing the use of machine learning and artificial intelligence in detecting and monitoring disease trends. The role of open collaboration between various stakeholders in modelling was stressed, advocating for the continuation of such partnerships beyond the pandemic. A major gap identified was the absence of a common international framework for data sharing, which is crucial for global pandemic preparedness. Overall, the workshop underscored the need for robust, adaptable modelling frameworks and the integration of different data sources and collaboration across sectors, as key elements in enhancing future pandemic response and preparedness.

Sepsis and underlying comorbidities in intensive care unit patients : Analysis of the cause of death by different clinicians-a pilot study.

BACKGROUND: There is an ongoing debate as to whether death with sepsis is primarily caused by sepsis or, more often, by the underlying disease. There are no data on the influence of a researcher's background on such an assessment. Therefore, the aim of this analysis was to assess the cause of death in sepsis and the influence of an investigator's professional background on such an assessment. MATERIALS AND METHODS: We performed a retrospective observational cohort study of sepsis patients treated in the medical intensive care unit (ICU) of a tertiary care center. For deceased patients, comorbidities and severity of illness were documented. The cause of death (sepsis or comorbidities or both combined) was independently assessed by four assessors with different professional backgrounds (medical student, senior physician in the medical ICU, anesthesiological intensivist, and senior physician specialized in the predominant comorbidity). RESULTS: In all, 78 of 235 patients died in hospital. Agreement between assessors about cause of death was low (κ 0.37, 95% confidence interval 0.29-0.44). Depending on the assessor, sepsis was the sole cause of death in 6-12% of cases, sepsis and comorbidities in 54-76%, and comorbidities alone in 18-40%. CONCLUSIONS: In a relevant proportion of patients with sepsis treated in the medical ICU, comorbidities contribute significantly to mortality, and death from sepsis without relevant comorbidities is a rare event. Designation of the cause of death in sepsis patients is highly subjective and may be influenced by the professional background of the assessor.

Integrating machine learning with pharmacokinetic models: Benefits of scientific machine learning in adding neural networks components to existing PK models.

Recently, the use of machine-learning (ML) models for pharmacokinetic (PK) modeling has grown significantly. Although most of the current approaches use ML techniques as black boxes, there are only a few that have proposed interpretable architectures which integrate mechanistic knowledge. In this work, we use as the test case a one-compartment PK model using a scientific machine learning (SciML) framework and consider learning an unknown absorption using neural networks, while simultaneously estimating other parameters of drug distribution and elimination. We generate simulated data with different sampling strategies to show that our model can accurately predict concentrations in extrapolation tasks, including new dosing regimens with different sparsity levels, and produce reliable forecasts even for new patients. By using a scenario of fitting PK data with complex absorption, we demonstrate that including known physiological structure into an SciML model allows us to obtain highly accurate predictions while preserving the interpretability of classical compartmental models.

Development of an early alert model for pandemic situations in Germany.

The COVID-19 pandemic has pointed out the need for new technical approaches to increase the preparedness of healthcare systems. One important measure is to develop innovative early warning systems. Along those lines, we first compiled a corpus of relevant COVID-19 related symptoms with the help of a disease ontology, text mining and statistical analysis. Subsequently, we applied statistical and machine learning (ML) techniques to time series data of symptom related Google searches and tweets spanning the time period from March 2020 to June 2022. In conclusion, we found that a long-short-term memory (LSTM) jointly trained on COVID-19 symptoms related Google Trends and Twitter data was able to accurately forecast up-trends in classical surveillance data (confirmed cases and hospitalization rates) 14 days ahead. In both cases, F1 scores were above 98% and 97%, respectively, hence demonstrating the potential of using digital traces for building an early alert system for pandemics in Germany.

Gene-based burden scores identify rare variant associations for 28 blood biomarkers.

BACKGROUND: A relevant part of the genetic architecture of complex traits is still unknown; despite the discovery of many disease-associated common variants. Polygenic risk score (PRS) models are based on the evaluation of the additive effects attributable to common variants and have been successfully implemented to assess the genetic susceptibility for many phenotypes. In contrast, burden tests are often used to identify an enrichment of rare deleterious variants in specific genes. Both kinds of genetic contributions are typically analyzed independently. Many studies suggest that complex phenotypes are influenced by both low effect common variants and high effect rare deleterious variants. The aim of this paper is to integrate the effect of both common and rare functional variants for a more comprehensive genetic risk modeling. METHODS: We developed a framework combining gene-based scores based on the enrichment of rare functionally relevant variants with genome-wide PRS based on common variants for association analysis and prediction models. We applied our framework on UK Biobank dataset with genotyping and exome data and considered 28 blood biomarkers levels as target phenotypes. For each biomarker, an association analysis was performed on full cohort using gene-based scores (GBS). The cohort was then split into 3 subsets for PRS construction and feature selection, predictive model training, and independent evaluation, respectively. Prediction models were generated including either PRS, GBS or both (combined). RESULTS: Association analyses of the cohort were able to detect significant genes that were previously known to be associated with different biomarkers. Interestingly, the analyses also revealed heterogeneous effect sizes and directionality highlighting the complexity of the blood biomarkers regulation. However, the combined models for many biomarkers show little or no improvement in prediction accuracy compared to the PRS models. CONCLUSION: This study shows that rare variants play an important role in the genetic architecture of complex multifactorial traits such as blood biomarkers. However, while rare deleterious variants play a strong role at an individual level, our results indicate that classical common variant based PRS might be more informative to predict the genetic susceptibility at the population level.

MultiGML: Multimodal graph machine learning for prediction of adverse drug events.

Adverse drug events constitute a major challenge for the success of clinical trials. Several computational strategies have been suggested to estimate the risk of adverse drug events in preclinical drug development. While these approaches have demonstrated high utility in practice, they are at the same time limited to specific information sources. Thus, many current computational approaches neglect a wealth of information which results from the integration of different data sources, such as biological protein function, gene expression, chemical compound structure, cell-based imaging and others. In this work we propose an integrative and explainable multi-modal Graph Machine Learning approach (MultiGML), which fuses knowledge graphs with multiple further data modalities to predict drug related adverse events and general drug target-phenotype associations. MultiGML demonstrates excellent prediction performance compared to alternative algorithms, including various traditional knowledge graph embedding techniques. MultiGML distinguishes itself from alternative techniques by providing in-depth explanations of model predictions, which point towards biological mechanisms associated with predictions of an adverse drug event. Hence, MultiGML could be a versatile tool to support decision making in preclinical drug development.

A Transformer-Based Model Trained on Large Scale Claims Data for Prediction of Severe COVID-19 Disease Progression.

In situations like the COVID-19 pandemic, healthcare systems are under enormous pressure as they can rapidly collapse under the burden of the crisis. Machine learning (ML) based risk models could lift the burden by identifying patients with a high risk of severe disease progression. Electronic Health Records (EHRs) provide crucial sources of information to develop these models because they rely on routinely collected healthcare data. However, EHR data is challenging for training ML models because it contains irregularly timestamped diagnosis, prescription, and procedure codes. For such data, transformer-based models are promising. We extended the previously published Med-BERT model by including age, sex, medications, quantitative clinical measures, and state information. After pre-training on approximately 988 million EHRs from 3.5 million patients, we developed models to predict Acute Respiratory Manifestations (ARM) risk using the medical history of 80,211 COVID-19 patients. Compared to Random Forests, XGBoost, and RETAIN, our transformer-based models more accurately forecast the risk of developing ARM after COVID-19 infection. We used Integrated Gradients and Bayesian networks to understand the link between the essential features of our model. Finally, we evaluated adapting our model to Austrian in-patient data. Our study highlights the promise of predictive transformer-based models for precision medicine.

A machine learning method for the identification and characterization of novel COVID-19 drug targets.

In addition to vaccines, the World Health Organization sees novel medications as an urgent matter to fight the ongoing COVID-19 pandemic. One possible strategy is to identify target proteins, for which a perturbation by an existing compound is likely to benefit COVID-19 patients. In order to contribute to this effort, we present GuiltyTargets-COVID-19 ( https://guiltytargets-covid.eu/ ), a machine learning supported web tool to identify novel candidate drug targets. Using six bulk and three single cell RNA-Seq datasets, together with a lung tissue specific protein-protein interaction network, we demonstrate that GuiltyTargets-COVID-19 is capable of (i) prioritizing meaningful target candidates and assessing their druggability, (ii) unraveling their linkage to known disease mechanisms, (iii) mapping ligands from the ChEMBL database to the identified targets, and (iv) pointing out potential side effects in the case that the mapped ligands correspond to approved drugs. Our example analyses identified 4 potential drug targets from the datasets: AKT3 from both the bulk and single cell RNA-Seq data as well as AKT2, MLKL, and MAPK11 in the single cell experiments. Altogether, we believe that our web tool will facilitate future target identification and drug development for COVID-19, notably in a cell type and tissue specific manner.

Bayesian network modeling of risk and prodromal markers of Parkinson's disease.

Parkinson's disease (PD) is characterized by a long prodromal phase with a multitude of markers indicating an increased PD risk prior to clinical diagnosis based on motor symptoms. Current PD prediction models do not consider interdependencies of single predictors, lack differentiation by subtypes of prodromal PD, and may be limited and potentially biased by confounding factors, unspecific assessment methods and restricted access to comprehensive marker data of prospective cohorts. We used prospective data of 18 established risk and prodromal markers of PD in 1178 healthy, PD-free individuals and 24 incident PD cases collected longitudinally in the Tübingen evaluation of Risk factors for Early detection of NeuroDegeneration (TREND) study at 4 visits over up to 10 years. We employed artificial intelligence (AI) to learn and quantify PD marker interdependencies via a Bayesian network (BN) with probabilistic confidence estimation using bootstrapping. The BN was employed to generate a synthetic cohort and individual marker profiles. Robust interdependencies were observed for BN edges from age to subthreshold parkinsonism and urinary dysfunction, sex to substantia nigra hyperechogenicity, depression, non-smoking and to constipation; depression to symptomatic hypotension and excessive daytime somnolence; solvent exposure to cognitive deficits and to physical inactivity; and non-smoking to physical inactivity. Conversion to PD was interdependent with prior subthreshold parkinsonism, sex and substantia nigra hyperechogenicity. Several additional interdependencies with lower probabilistic confidence were identified. Synthetic subjects generated via the BN based representation of the TREND study were realistic as assessed through multiple comparison approaches of real and synthetic data. Altogether our work demonstrates the potential of modern AI approaches (specifically BNs) both for modelling and understanding interdependencies between PD risk and prodromal markers, which are so far not accounted for in PD prediction models, as well as for generating realistic synthetic data.

AI reveals insights into link between CD33 and cognitive impairment in Alzheimer's Disease.

Modeling biological mechanisms is a key for disease understanding and drug-target identification. However, formulating quantitative models in the field of Alzheimer's Disease is challenged by a lack of detailed knowledge of relevant biochemical processes. Additionally, fitting differential equation systems usually requires time resolved data and the possibility to perform intervention experiments, which is difficult in neurological disorders. This work addresses these challenges by employing the recently published Variational Autoencoder Modular Bayesian Networks (VAMBN) method, which we here trained on combined clinical and patient level gene expression data while incorporating a disease focused knowledge graph. Our approach, called iVAMBN, resulted in a quantitative model that allowed us to simulate a down-expression of the putative drug target CD33, including potential impact on cognitive impairment and brain pathophysiology. Experimental validation demonstrated a high overlap of molecular mechanism predicted to be altered by CD33 perturbation with cell line data. Altogether, our modeling approach may help to select promising drug targets.

Artificial intelligence-based clustering and characterization of Parkinson's disease trajectories.

Parkinson's disease (PD) is a highly heterogeneous disease both with respect to arising symptoms and its progression over time. This hampers the design of disease modifying trials for PD as treatments which would potentially show efficacy in specific patient subgroups could be considered ineffective in a heterogeneous trial cohort. Establishing clusters of PD patients based on their progression patterns could help to disentangle the exhibited heterogeneity, highlight clinical differences among patient subgroups, and identify the biological pathways and molecular players which underlie the evident differences. Further, stratification of patients into clusters with distinct progression patterns could help to recruit more homogeneous trial cohorts. In the present work, we applied an artificial intelligence-based algorithm to model and cluster longitudinal PD progression trajectories from the Parkinson's Progression Markers Initiative. Using a combination of six clinical outcome scores covering both motor and non-motor symptoms, we were able to identify specific clusters of PD that showed significantly different patterns of PD progression. The inclusion of genetic variants and biomarker data allowed us to associate the established progression clusters with distinct biological mechanisms, such as perturbations in vesicle transport or neuroprotection. Furthermore, we found that patients of identified progression clusters showed significant differences in their responsiveness to symptomatic treatment. Taken together, our work contributes to a better understanding of the heterogeneity encountered when examining and treating patients with PD, and points towards potential biological pathways and genes that could underlie those differences.

A Sequential Targeting Strategy Interrupts AKT-Driven Subclone-Mediated Progression in Glioblastoma.

PURPOSE: Therapy resistance and fatal disease progression in glioblastoma are thought to result from the dynamics of intra-tumor heterogeneity. This study aimed at identifying and molecularly targeting tumor cells that can survive, adapt, and subclonally expand under primary therapy. EXPERIMENTAL DESIGN: To identify candidate markers and to experimentally access dynamics of subclonal progression in glioblastoma, we established a discovery cohort of paired vital cell samples obtained before and after primary therapy. We further used two independent validation cohorts of paired clinical tissues to test our findings. Follow-up preclinical treatment strategies were evaluated in patient-derived xenografts. RESULTS: We describe, in clinical samples, an archetype of rare ALDH1A1+ tumor cells that enrich and acquire AKT-mediated drug resistance in response to standard-of-care temozolomide (TMZ). Importantly, we observe that drug resistance of ALDH1A1+ cells is not intrinsic, but rather an adaptive mechanism emerging exclusively after TMZ treatment. In patient cells and xenograft models of disease, we recapitulate the enrichment of ALDH1A1+ cells under the influence of TMZ. We demonstrate that their subclonal progression is AKT-driven and can be interfered with by well-timed sequential rather than simultaneous antitumor combination strategy. CONCLUSIONS: Drug-resistant ALDH1A1+/pAKT+ subclones accumulate in patient tissues upon adaptation to TMZ therapy. These subclones may therefore represent a dynamic target in glioblastoma. Our study proposes the combination of TMZ and AKT inhibitors in a sequential treatment schedule as a rationale for future clinical investigation.

Modeling approaches for early warning and monitoring of pandemic situations as well as decision support.

The COVID-19 pandemic has highlighted the lack of preparedness of many healthcare systems against pandemic situations. In response, many population-level computational modeling approaches have been proposed for predicting outbreaks, spatiotemporally forecasting disease spread, and assessing as well as predicting the effectiveness of (non-) pharmaceutical interventions. However, in several countries, these modeling efforts have only limited impact on governmental decision-making so far. In light of this situation, the review aims to provide a critical review of existing modeling approaches and to discuss the potential for future developments.

Critical assessment of transformer-based AI models for German clinical notes.

Objective: Healthcare data such as clinical notes are primarily recorded in an unstructured manner. If adequately translated into structured data, they can be utilized for health economics and set the groundwork for better individualized patient care. To structure clinical notes, deep-learning methods, particularly transformer-based models like Bidirectional Encoder Representations from Transformers (BERT), have recently received much attention. Currently, biomedical applications are primarily focused on the English language. While general-purpose German-language models such as GermanBERT and GottBERT have been published, adaptations for biomedical data are unavailable. This study evaluated the suitability of existing and novel transformer-based models for the German biomedical and clinical domain. Materials and Methods: We used 8 transformer-based models and pre-trained 3 new models on a newly generated biomedical corpus, and systematically compared them with each other. We annotated a new dataset of clinical notes and used it with 4 other corpora (BRONCO150, CLEF eHealth 2019 Task 1, GGPONC, and JSynCC) to perform named entity recognition (NER) and document classification tasks. Results: General-purpose language models can be used effectively for biomedical and clinical natural language processing (NLP) tasks, still, our newly trained BioGottBERT model outperformed GottBERT on both clinical NER tasks. However, training new biomedical models from scratch proved ineffective. Discussion: The domain-adaptation strategy's potential is currently limited due to a lack of pre-training data. Since general-purpose language models are only marginally inferior to domain-specific models, both options are suitable for developing German-language biomedical applications. Conclusion: General-purpose language models perform remarkably well on biomedical and clinical NLP tasks. If larger corpora become available in the future, domain-adapting these models may improve performances.