Benjamin Wild

Benjamin Wild

Auto-generated profile, based on publicly available information

Benjamin Wild

Center of Digital Health, Berlin Institute of Health at Charité – Universitätsmedizin Berlin, Berlin, Germany

Research Focus & Metrics

Academic Career Length

12Years (2015-Present)

(Calculated once Scilit publication data is synced: current year − year of first publication + 1)

Representative Publications

View All Publications
22 September 2022 in Nature Medicine
Verified
Journal Article
Open access
PDF file
TBThore Buergel
Thore Buergel
JSJakob Steinfeldt
Jakob Steinfeldt
GRGreg Ruyoga
Greg Ruyoga
MPMaik Pietzner
Maik Pietzner
DBDaniele Bizzarri
Daniele Bizzarri
ULUlf Landmesser
Ulf Landmesser
Risk stratification is critical for the early identification of high-risk individuals and disease prevention. Here we explored the potential of nuclear magnetic resonance (NMR) spectroscopy-derived metabolomic profiles to inform on multidisease risk beyond conventional clinical predictors for the onset of 24 common conditions, including metabolic, vascular, respiratory, musculoskeletal and neurological diseases and cancers. Specifically, we trained a neural network to learn disease-specific metabolomic states from 168 circulating metabolic markers measured in 117,981 participants with ~1.4 million person-years of follow-up from the UK Biobank and validated the model in four independent cohorts. We found metabolomic states to be associated with incident event rates in all the investigated conditions, except breast cancer. For 10-year outcome prediction for 15 endpoints, with and without established metabolic contribution, a combination of age and sex and the metabolomic state equaled or outperformed established predictors. Moreover, metabolomic state added predictive information over comprehensive clinical variables for eight common diseases, including type 2 diabetes, dementia and heart failure. Decision curve analyses showed that predictive improvements translated into clinical utility for a wide range of potential decision thresholds. Taken together, our study demonstrates both the potential and limitations of NMR-derived metabolomic profiles as a multidisease assay to inform on the risk of many common diseases simultaneously.

23 April 2025 in Nature Medicine
Verified
Journal Article
Open access
PDF file
SSSarah Sandmann
Sarah Sandmann
SHStefan Hegselmann
Stefan Hegselmann
MFMichael Fujarski
Michael Fujarski
LBLucas Bickmann
Lucas Bickmann
BWBenjamin Wild
Benjamin Wild
Julian VargheseJulian Varghese
Julian Varghese
Large Language Models (LLMs) are increasingly transforming medical applications. However, proprietary models such as GPT-4o face significant barriers to clinical adoption because they cannot be deployed on site within healthcare institutions, making them non-compliant with stringent privacy regulations. Recent advancements in open-source LLMs such as DeepSeek models offer a promising alternative since they allow efficient fine-tuning on local data in hospitals with advanced IT infrastructure. To demonstrate the clinical utility of DeepSeek-V3 and DeepSeek-R1, we benchmarked their performance on clinical decision support tasks against proprietary LLMs, including GPT-4o and Gemini-2.0 Flash Thinking Experimental. Using 125 patient cases with sufficient statistical power, covering a broad range of frequent and rare diseases, we found that DeepSeek models perform equally well and in some cases better than proprietary LLMs. Our study demonstrates that open-source LLMs can provide a scalable pathway for secure model training enabling real-world medical applications in accordance with data privacy and healthcare regulations.

17 February 2021 in Nature Communications
Verified
Journal Article
Open access
PDF file
BWBenjamin Wild
Benjamin Wild
DDDavid M. Dormagen
David M. Dormagen
AZAdrian Zachariae
Adrian Zachariae
MSMichael L. Smith
Michael L. Smith
Kirsten TraynorKirsten Traynor
Kirsten Traynor
TLTim Landgraf
Tim Landgraf
In complex societies, individuals’ roles are reflected by interactions with other conspecifics. Honey bees (Apis mellifera) generally change tasks as they age, but developmental trajectories of individuals can vary drastically due to physiological and environmental factors. We introduce a succinct descriptor of an individual’s social network that can be obtained without interfering with the colony. This ‘network age’ accurately predicts task allocation, survival, activity patterns, and future behavior. We analyze developmental trajectories of multiple cohorts of individuals in a natural setting and identify distinct developmental pathways and critical life changes. Our findings suggest a high stability in task allocation on an individual level. We show that our method is versatile and can extract different properties from social networks, opening up a broad range of future studies. Our approach highlights the relationship of social interactions and individual traits, and provides a scalable technique for understanding how complex social systems function.

25 September 2015 in Frontiers in Ecology and Evolution
Verified
Journal Article
Open access
PDF file
FWFernando Wario
Fernando Wario
BWBenjamin Wild
Benjamin Wild
MCMargaret J. Couvillon
Margaret J. Couvillon
RRRaúl Rojas
Raúl Rojas
TLTim Landgraf
Tim Landgraf
The honeybee waggle dance communication system is an intriguing example of abstract animal communication and has been investigated thoroughly throughout the last seven decades. Typically, observables such as durations or angles are extracted manually directly from the observation hive or from video recordings to quantify dance properties, particularly to determine where bees have foraged. In recent years, biology has profited from automation, improving measurement precision, removing human bias, and accelerating data collection. As a further step, we have developed technologies to track all individuals of a honeybee colony and detect and decode communication dances automatically. In strong contrast to conventional approaches that focus on a small subset of the hive life, whether this regards time, space, or animal identity, our more inclusive system will help the understanding of the dance comprehensively in its spatial, temporal, and social context. In this contribution, we present full specifications of the recording setup and the software for automatic recognition and decoding of tags and dances, and we discuss potential research directions that may benefit from automation. Lastly, to exemplify the power of the methodology, we show experimental data and respective analyses for a continuous, experimental recording of nine weeks duration.

18 October 2023 in Journal of The Royal Society Interface
Verified
Journal Article
Open access
JJJitesh Jhawar
Jitesh Jhawar
JDJacob D. Davidson
Jacob D. Davidson
AWAnja Weidenmüller
Anja Weidenmüller
BWBenjamin Wild
Benjamin Wild
DDDavid M. Dormagen
David M. Dormagen
MSMichael L. Smith
Michael L. Smith
A honey bee colony functions as an integrated collective, with individuals coordinating their behaviour to adapt and respond to unexpected disturbances. Nest homeostasis is critical for colony function; when ambient temperatures increase, individuals switch to thermoregulatory roles to cool the nest, such as fanning and water collection. While prior work has focused on bees engaged in specific behaviours, less is known about how responses are coordinated at the colony level, and how previous tasks predict behavioural changes during a heat stress. Using BeesBook automated tracking, we follow thousands of individuals during an experimentally induced heat stress, and analyse their behavioural changes from the individual to colony level. We show that heat stress causes an overall increase in activity levels and a spatial reorganization of bees away from the brood area. Using a generalized framework to analyse individual behaviour, we find that individuals differ in their response to heat stress, which depends on their prior behaviour and correlates with age. Examining the correlation of behavioural metrics over time suggests that heat stress perturbation does not have a long-lasting effect on an individual’s future behaviour. These results demonstrate how thousands of individuals within a colony change their behaviour to achieve a coordinated response to an environmental disturbance.

22 September 2025 in European Journal of Heart Failure
Verified
Journal Article
Open access
JVJakob Versnjak
Jakob Versnjak
TKTitus Kuehne
Titus Kuehne
PFPauline Fahjen
Pauline Fahjen
NJNina Jovanovic
Nina Jovanovic
ULUlrike Löber
Ulrike Löber
MKMarcus Kelm
Marcus Kelm
Aims: Heart failure with preserved ejection fraction (HFpEF) has become the predominant form of heart failure and a leading cause of global cardiovascular morbidity and mortality. Due to its heterogeneous nature, HFpEF presents substantial challenges in diagnosis and management. Given the limited treatment options and lifestyle-associated comorbidities, early identification is crucial for establishing effective preventive strategies. Here, we introduce and validate a machine learning-based multi-omics approach that integrates clinical and molecular data to detect and characterize HFpEF. Methods and results: A supervised classifier was trained on a stratified subset of UK Biobank participants (n = 401 917) to identify phenotypic profiles associated with subsequent symptom-defined HFpEF during longitudinal follow-up. Model performance was validated in a non-overlapping hold-out subset from all 22 UK Biobank assessment centres (n = 100 446; 6726 HFpEF cases; 7394 with multi-omics data). The classifier demonstrated robust discriminatory performance, with a receiver operating characteristic area under the curve (ROC AUC) of 0.931 (95% confidence interval [CI] 0.930–0.931), a sensitivity of 0.857 (95% CI 0.855–0.860) and a specificity of 0.847 (95% CI 0.846–0.847). It identified individuals who subsequently developed HFpEF an average of 6.3 ± 3.9 years before symptom onset in asymptomatic individuals. Similarity network fusion (SNF) identified distinct subgroups, including a high-risk cluster characterized by elevated mortality and dysregulated inflammatory pathways, which was distinguishable with high accuracy (ROC AUC 0.988; 95% CI 0.985–0.990). Conclusions: We identified HFpEF phenotypes at an early stage, often several years before the onset of clinical symptoms, when the disease trajectory may still be amenable to modification. The molecular characterization provides novel insights into the underlying disease complexity and enables more refined risk stratification.