BEGIN:VCALENDAR
VERSION:2.0
PRODID:-//Biomedical Mathematics Group - ECPv6.17.2//NONSGML v1.0//EN
CALSCALE:GREGORIAN
METHOD:PUBLISH
X-ORIGINAL-URL:https://www.ibs.re.kr/bimag
X-WR-CALDESC:Events for Biomedical Mathematics Group
REFRESH-INTERVAL;VALUE=DURATION:PT1H
X-Robots-Tag:noindex
X-PUBLISHED-TTL:PT1H
BEGIN:VTIMEZONE
TZID:Asia/Seoul
BEGIN:STANDARD
TZOFFSETFROM:+0900
TZOFFSETTO:+0900
TZNAME:KST
DTSTART:20250101T000000
END:STANDARD
END:VTIMEZONE
BEGIN:VEVENT
DTSTART;TZID=Asia/Seoul:20260410T110000
DTEND;TZID=Asia/Seoul:20260410T120000
DTSTAMP:20260311T121617Z
CREATED:20260205T074338Z
LAST-MODIFIED:20260311T121617Z
UID:12182-1775818800-1775822400@www.ibs.re.kr
SUMMARY:A Data-Driven Computational Framework for Identifiability and Nonlinear Dynamics Discovery in Complex Systems - Wenrui Hao
DESCRIPTION:Data-driven modeling is essential for deciphering complex biological systems\, yet its utility is often constrained by two fundamental hurdles: the inability to guarantee parameter identifiability and the high computational cost of learning nonlinear dynamics. This talk introduces a unified computational framework designed to overcome these challenges\, bridging theoretical rigor with scalable machine learning. \n\nThe first component of the framework establishes a computational foundation for practical identifiability. By leveraging the Fisher Information Matrix and its theoretical links to coordinate identifiability\, we propose an efficient method for identifiability assessment. We further introduce regularization-based strategies to manage non-identifiable parameters\, thereby enhancing model reliability and facilitating robust uncertainty quantification. \n\nTo address the discovery of nonlinear dynamics\, we present the Laplacian Eigenfunction-Based Neural Operator (LE-NO). This operator learning framework is specifically engineered for modeling reaction–diffusion equations. By projecting nonlinear operators onto Laplacian eigenfunctions\, LE-NO achieves superior computational efficiency and generalization across varying boundary conditions\, effectively bypassing the limitations of large-scale architectures and data scarcity. \n\nFinally\, we demonstrate the framework’s utility in the context of Alzheimer’s disease modeling. We show that this integrated approach ensures reliable parameter inference while capturing the intricate nonlinear dynamics of disease progression\, providing a critical step toward the development of high-fidelity digital twins for neurodegenerative pathology. \n  \nZoom : 997 8258 4700 (pw : 1234)
URL:https://www.ibs.re.kr/bimag/event/a-data-driven-computational-framework-for-identifiability-and-nonlinear-dynamics-discovery-in-complex-systems-wenrui-hao/
LOCATION:B232 Seminar Room\, IBS\, 55 Expo-ro Yuseong-gu\, Daejeon\, Daejeon\, 34126\, Korea\, Republic of
CATEGORIES:Biomedical Mathematics Online Colloquium
ATTACH;FMTTYPE=image/jpeg:https://www.ibs.re.kr/bimag/cms/wp-content/uploads/2026/02/Wenrui-Hao-2-e1770278378786.jpg
ORGANIZER;CN="Jae Kyoung Kim":MAILTO:jaekkim@kaist.ac.kr
END:VEVENT
BEGIN:VEVENT
DTSTART;TZID=Asia/Seoul:20260417T100000
DTEND;TZID=Asia/Seoul:20260417T120000
DTSTAMP:20260406T060603Z
CREATED:20260403T080037Z
LAST-MODIFIED:20260406T060603Z
UID:12336-1776420000-1776427200@www.ibs.re.kr
SUMMARY:Discovering network dynamics with neural symbolic regression - Olive Cawiding
DESCRIPTION:In this tallk\, we discuss the paper “Discovering network dynamics with neural symbolic regression” by Zihan Yu et al.\, Nature Com. Science\, 2026. \nAbstract  \nNetwork dynamics are fundamental to analyzing the properties of high-dimensional complex systems and understanding their behavior. Despite the accumulation of observational data across many domains\, mathematical models exist in only a few areas with clear underlying principles. Here we show that a neural symbolic regression approach can bridge this gap by automatically deriving formulas from data. Our method reduces searches on high-dimensional networks to equivalent one-dimensional systems and uses pretrained neural networks to guide accurate formula discovery. Applied to ten benchmark systems\, it recovers the correct forms and parameters of underlying dynamics. In two empirical natural systems\, it corrects existing models of gene regulation and microbial communities\, reducing prediction error by 59.98% and 55.94%\, respectively. In epidemic transmission across human mobility networks of various scales\, it discovers dynamics that exhibit the same power-law distribution of node correlations across scales and reveal country-level differences in intervention effects. These results demonstrate that machine-driven discovery of network dynamics can enhance understandings of complex systems and advance the development of complexity science.
URL:https://www.ibs.re.kr/bimag/event/discovering-network-dynamics-with-neural-symbolic-regression-olive-cawiding/
LOCATION:B232 Seminar Room\, IBS\, 55 Expo-ro Yuseong-gu\, Daejeon\, Daejeon\, 34126\, Korea\, Republic of
CATEGORIES:Journal Club
ORGANIZER;CN="Jae Kyoung Kim":MAILTO:jaekkim@kaist.ac.kr
END:VEVENT
BEGIN:VEVENT
DTSTART;TZID=Asia/Seoul:20260424T100000
DTEND;TZID=Asia/Seoul:20260424T120000
DTSTAMP:20260406T103749Z
CREATED:20260406T092550Z
LAST-MODIFIED:20260406T103749Z
UID:12365-1777024800-1777032000@www.ibs.re.kr
SUMMARY:Foundation Models for Wearable Movement Data in Mental Health Research - Aqsa Awan
DESCRIPTION:In this tallk\, we discuss the paper “Foundation Models for Wearable Movement Data in Mental Health Research” by Franklin Y. Ruan et al.\, arXiv\, 2025. \nAbstract \nPretrained foundation models and transformer architectures have driven the success of large language models (LLMs) and other modern AI breakthroughs. However\, similar advancements in health data modeling remain limited due to the need for innovative adaptations. Wearable movement data offers a valuable avenue for exploration\, as it’s a core feature in nearly all commercial smartwatches\, well established in clinical and mental health research\, and the sequential nature of the data shares similarities to language. We introduce the Pretrained Actigraphy Transformer (PAT)\, the first open source foundation model designed for time-series wearable movement data. Leveraging transformer-based architectures and novel techniques\, such as patch embeddings\, and pretraining on data from 29\,307 participants in a national U.S. sample\, PAT achieves state-of-the-art performance in several mental health prediction tasks. PAT is also lightweight and easily interpretable\, making it a robust tool for mental health research.
URL:https://www.ibs.re.kr/bimag/event/foundation-models-for-wearable-movement-data-in-mental-health-research-aqsa-awan/
LOCATION:B232 Seminar Room\, IBS\, 55 Expo-ro Yuseong-gu\, Daejeon\, Daejeon\, 34126\, Korea\, Republic of
CATEGORIES:Journal Club
ORGANIZER;CN="Jae Kyoung Kim":MAILTO:jaekkim@kaist.ac.kr
END:VEVENT
BEGIN:VEVENT
DTSTART;TZID=Asia/Seoul:20260506T160000
DTEND;TZID=Asia/Seoul:20260506T170000
DTSTAMP:20260311T121631Z
CREATED:20260205T074722Z
LAST-MODIFIED:20260311T121631Z
UID:12189-1778083200-1778086800@www.ibs.re.kr
SUMMARY:Data-driven discovery of biological oscillator models - Lendert Gelens
DESCRIPTION:Oscillatory dynamics are a found everywhere in living systems\, underlying processes such as metabolic regulation\, cell division\, and embryonic development. Identifying the mechanisms that generate these rhythms is challenging due to nonlinear interactions\, multiple time scales\, and limited access to all relevant variables. Data-driven approaches offer a promising route to infer dynamical models directly from time-series data. In this talk\, I will discuss our work on data-driven discovery of models for (bio)chemical oscillators. In particular\, I will present CLINE\, a neural-network–based framework that infers key geometric features of phase space\, such as nullclines\, from oscillatory data and uses this information to construct low-dimensional dynamical models. \n  \nZoom : 997 8258 4700 (pw : 1234)
URL:https://www.ibs.re.kr/bimag/event/data-driven-discovery-of-biological-oscillator-models-lendert-gelens/
LOCATION:B232 Seminar Room\, IBS\, 55 Expo-ro Yuseong-gu\, Daejeon\, Daejeon\, 34126\, Korea\, Republic of
CATEGORIES:Biomedical Mathematics Online Colloquium
ATTACH;FMTTYPE=image/jpeg:https://www.ibs.re.kr/bimag/cms/wp-content/uploads/2026/02/Gelens_Lendert_alumni-e1770278337776.jpg
ORGANIZER;CN="Jae Kyoung Kim":MAILTO:jaekkim@kaist.ac.kr
END:VEVENT
BEGIN:VEVENT
DTSTART;TZID=Asia/Seoul:20260508T100000
DTEND;TZID=Asia/Seoul:20260508T120000
DTSTAMP:20260506T033947Z
CREATED:20260406T041825Z
LAST-MODIFIED:20260506T033947Z
UID:12360-1778234400-1778241600@www.ibs.re.kr
SUMMARY:Impact of daylight saving time on physical activity patterns - Myna Lim
DESCRIPTION:In this talk\, we discuss the paper “Impact of daylight saving time on physical activity patterns” by Hayoung Jeong et al.\, Nature Health\, 2026. \nAbstract\nDaylight saving time (DST) remains contentious: some policymakers highlight behavioural benefits\, while others emphasize health risks. Here we estimated the behavioural and physiological impacts of DST using longitudinal Fitbit measures from the National Institutes of Health All of Us Research Program. Avoiding strict modelling assumptions\, we used a natural difference-in-differences design with Arizona (no DST) as a control against neighbouring Mountain Time states (observing DST). Contrary to common belief\, DST transitions produced no net change in total daily steps. Instead\, activity was reallocated to other times of day: fall transitions increased morning steps by 202 (confidence interval = [78\, 326]\, P = 0.001) while reducing evening steps by 180 (confidence interval = [−263\, −97]\, P < 0.001); spring transitions showed the opposite. Importantly\, these treatment effects varied by demographics and across data-driven activity phenotypes (‘morning walker’\, ‘neutral walker’ and ‘evening walker’). These disparities suggest that structural factors (for example\, rigid work schedules\, perceived safety) may constrain the capacity to flexibly adapt to time shifts for some populations. Physiologically\, resting heart rate showed subtle intraday shifts mirroring behavioural changes\, although differences were clinically insignificant. Our study provides a large-scale causal analysis of DST’s influence using continuous wearables data\, illustrating how observational data can generate real-world evidence to inform health-relevant policies.
URL:https://www.ibs.re.kr/bimag/event/digital-biomarkers-for-brain-health-passive-and-continuous-assessment-from-wearable-sensors-myna-lim/
LOCATION:B232 Seminar Room\, IBS\, 55 Expo-ro Yuseong-gu\, Daejeon\, Daejeon\, 34126\, Korea\, Republic of
CATEGORIES:Journal Club
ORGANIZER;CN="Jae Kyoung Kim":MAILTO:jaekkim@kaist.ac.kr
END:VEVENT
BEGIN:VEVENT
DTSTART;TZID=Asia/Seoul:20260515T100000
DTEND;TZID=Asia/Seoul:20260515T120000
DTSTAMP:20260429T070938Z
CREATED:20260403T080250Z
LAST-MODIFIED:20260429T070938Z
UID:12338-1778839200-1778846400@www.ibs.re.kr
SUMMARY:High-order Michaelis-Menten equations allow inference of hidden kinetic parameters in enzyme catalysis - Hyeong Jun Jang
DESCRIPTION:In this talk\, we discuss the paper “High-order Michaelis-Menten equations allow inference of hidden kinetic parameters in enzyme catalysis” by Divya Singh et al.\, Nat. Comm.\, 2025. \nAbstract \nSingle-molecule measurements provide a platform for investigating the dynamical properties of enzymatic reactions. To this end\, the single-molecule Michaelis-Menten equation was instrumental as it asserts that the first moment of the enzymatic turnover time depends linearly on the reciprocal of the substrate concentration. This\, in turn\, provides robust and convenient means to determine the maximal turnover rate and the Michaelis-Menten constant. Yet\, the information provided by these parameters is incomplete and does not allow access to key observables such as the lifetime of the enzyme-substrate complex\, the rate of substrate-enzyme binding\, and the probability of successful product formation. Here we show that these quantities and others can be inferred via a set of high-order Michaelis-Menten equations that we derive. These equations capture universal linear relations between the reciprocal of the substrate concentration and distinguished combinations of turnover time moments\, essentially generalizing the Michaelis-Menten equation to moments of any order. We demonstrate how key observables such as the lifetime of the enzyme-substrate complex\, the rate of substrate-enzyme binding\, and the probability of successful product formation\, can all be inferred using these high-order Michaelis-Menten equations. We test our inference procedure to show that it is robust\, producing accurate results with only several thousand turnover events per substrate concentration.
URL:https://www.ibs.re.kr/bimag/event/high-order-michaelis-menten-equations-allow-inference-of-hidden-kinetic-parameters-in-enzyme-catalysis-hyeong-jun-jang/
LOCATION:B232 Seminar Room\, IBS\, 55 Expo-ro Yuseong-gu\, Daejeon\, Daejeon\, 34126\, Korea\, Republic of
CATEGORIES:Journal Club
ORGANIZER;CN="Jae Kyoung Kim":MAILTO:jaekkim@kaist.ac.kr
END:VEVENT
BEGIN:VEVENT
DTSTART;TZID=Asia/Seoul:20260518T123000
DTEND;TZID=Asia/Seoul:20260518T133000
DTSTAMP:20260427T133007Z
CREATED:20260427T132625Z
LAST-MODIFIED:20260427T133007Z
UID:12390-1779107400-1779111000@www.ibs.re.kr
SUMMARY:Heejung Shim - Modelling spatial transcriptomics: from flexible cell-type deconvolution to multi-scale spatial factor analysis
DESCRIPTION:Abstract: \nSpatial transcriptomics enables the study of gene expression within its spatial context\, but introduces key statistical challenges\, including mixed cellular composition and complex spatial structure. In this talk\, I present two complementary modelling approaches.First\, I introduce FlexiDeconv\, a cell-type deconvolution method based on a modified Latent Dirichlet Allocation framework. A key feature of this method is its flexible use of reference information\, allowing the model to balance prior information from scRNA-seq with signals from observed spatial data\, and to adapt when the reference is incomplete or mismatched\, a common challenge in practice.I then present WaviFM\, a wavelet-based Bayesian sparse factor model that captures spatial gene expression patterns across multiple spatial scales\, enabling the detection of both fine and broad spatial patterns. In addition\, WaviFM can incorporate gene-set information to guide factor inference\, while allowing for uncertainty and potential errors in these annotations.Together\, these methods illustrate how flexible modelling of prior information and multi-scale modelling of spatial structure can improve our ability to extract biologically meaningful signals from spatial transcriptomics data.
URL:https://www.ibs.re.kr/bimag/event/heejung-shim-modelling-spatial-transcriptomics-from-flexible-cell-type-deconvolution-to-multi-scale-spatial-factor-analysis/
LOCATION:B232 Seminar Room\, IBS\, 55 Expo-ro Yuseong-gu\, Daejeon\, Daejeon\, 34126\, Korea\, Republic of
CATEGORIES:Biomedical Mathematics Seminar
ORGANIZER;CN="Jae Kyoung Kim":MAILTO:jaekkim@kaist.ac.kr
END:VEVENT
BEGIN:VEVENT
DTSTART;TZID=Asia/Seoul:20260522T110000
DTEND;TZID=Asia/Seoul:20260522T120000
DTSTAMP:20260311T121645Z
CREATED:20260205T075139Z
LAST-MODIFIED:20260311T121645Z
UID:12194-1779447600-1779451200@www.ibs.re.kr
SUMMARY:Mathematics of diffusive signaling - Alan Lindsay
DESCRIPTION:Diffusive transport is one of the most fundamental mechanisms by which information\, mass\, and chemical signals propagate in physical and biological systems. In many settings—ranging from cellular signaling to chemical sensing—communication is mediated by particles undergoing random motion and interacting with small\, spatially localized targets. This talk explores the mathematical structures underlying diffusive signaling\, emphasizing how geometry\, stochasticity\, and multiscale effects shape signal detection and reliability. Using tools from stochastic processes\, partial differential equations\, and asymptotic analysis\, I will describe how seemingly microscopic features can exert a dominant influence on macroscopic signaling outcomes\, and highlight recent progress on quantifying signal strength\, timing\, and variability in complex geometries. \n  \nZoom : 997 8258 4700 (pw : 1234)
URL:https://www.ibs.re.kr/bimag/event/mathematics-of-diffusive-signaling-alan-lindsay/
LOCATION:B232 Seminar Room\, IBS\, 55 Expo-ro Yuseong-gu\, Daejeon\, Daejeon\, 34126\, Korea\, Republic of
CATEGORIES:Biomedical Mathematics Online Colloquium
ATTACH;FMTTYPE=image/jpeg:https://www.ibs.re.kr/bimag/cms/wp-content/uploads/2026/02/alan_lindsay-e1770278281837.jpg
ORGANIZER;CN="Jae Kyoung Kim":MAILTO:jaekkim@kaist.ac.kr
END:VEVENT
BEGIN:VEVENT
DTSTART;TZID=Asia/Seoul:20260527T160000
DTEND;TZID=Asia/Seoul:20260527T170000
DTSTAMP:20260523T023337Z
CREATED:20260523T013451Z
LAST-MODIFIED:20260523T023337Z
UID:12489-1779897600-1779901200@www.ibs.re.kr
SUMMARY:Causal Generalist Medical AI - Hongtu Zhu
DESCRIPTION:The rapid evolution of flexible and reusable artificial intelligence (AI) models is transforming medical science. This short course introduces Causal Generalist Medical AI (Causal GMAI)—a paradigm that integrates causal inference with generalist AI models to enhance interpretability\, robustness\, and generalizability in medical decision-making. Causal GMAI employs self-supervised\, semi-supervised\, and supervised learning on diverse multimodal datasets—including imaging\, electronic health records\, clinical trials\,  laboratory results\, genomics\, knowledge graphs\, and medical text—to perform a wide range of tasks with minimal task-specific supervision.  By embedding causal reasoning\, these models go beyond prediction to infer underlying causal relationships\, improving diagnostic accuracy\, treatment recommendations\, and personalized medicine. The course covers key technical components such as causal discovery\, counterfactual reasoning\, and domain adaptation\, alongside real-world applications.  We will also explore challenges in regulation\, validation\, and dataset curation to ensure clinical reliability and ethical deployment. Designed for researchers\, clinicians\, data scientists\, and AI practitioners\, this course provides a foundation for advancing the next generation of trustworthy and interpretable medical AI. \n  \nZoom : 997 8258 4700 (pw : 1234)
URL:https://www.ibs.re.kr/bimag/event/hongtu-zhu-tba/
LOCATION:ZOOM ID: 997 8258 4700 (Biomedical Mathematics Online Colloquium)\, (pw: 1234)
CATEGORIES:Biomedical Mathematics Online Colloquium
ATTACH;FMTTYPE=image/png:https://www.ibs.re.kr/bimag/cms/wp-content/uploads/2026/05/hongtu.png
ORGANIZER;CN="Jae Kyoung Kim":MAILTO:jaekkim@kaist.ac.kr
END:VEVENT
BEGIN:VEVENT
DTSTART;TZID=Asia/Seoul:20260529T100000
DTEND;TZID=Asia/Seoul:20260529T120000
DTSTAMP:20260518T051101Z
CREATED:20260429T070610Z
LAST-MODIFIED:20260518T051101Z
UID:12398-1780048800-1780056000@www.ibs.re.kr
SUMMARY:Circadian rhythm profiles derived from accelerometer measures of the sleep-wake cycle in two cohort studies - Chitaranjan Mahapatra
DESCRIPTION:In this talk\, we discuss the paper “Circadian rhythm profiles derived from accelerometer measures of the sleep-wake cycle in two cohort studies” by Sam vidil et al.\, Nature Communications\, 2025. \nAbstract: \nAccelerometers allow objective measures of dimensions (rest-activity rhythm (RAR)\, daytime activity\, sleep\, and chronotype) of the bio-behavioural manifestation of circadian rhythm (CR) using multiple metrics in large-scale studies. These dimensions are rarely examined together due to methodological challenges of using correlated data. To address this challenge\, we propose a two-step approach consisting of data reduction of CR metrics using principal component analyses\, followed by k-means clustering to identify groups of individuals with a similar profile using data from the Whitehall II (N = 3\,991\, mean age=69.4years) and UK Biobank (N = 54\,995\, mean age=67.5years) cohort studies. Our analyses identified nine CR clusters: two presented extreme (most robust/poorest) RAR and (highest/lowest) daytime activity\, two robust RAR with opposite sleep profiles (longer and efficient/shorter and fragmented)\, one high-intensity physical activity\, and four poor RAR (one characterised by late chronotype\, two by low activity but opposite sleep profiles\, and one by restless (agitated) sleep). The participants in these nine clusters differed on sociodemographic\, behavioural and health-related factors. Findings were similar in these two independent cohort studies\, highlighting the validity of our approach. Most previous studies have used only the RAR dimension of circadian rhythm\, and here we show that this might be an oversimplification as demonstrated by nine clusters characterised by combinations of RAR\, daytime activity\, sleep\, and chronotype. Our innovative approach demonstrates feasibility of using all dimensions to study the impact of circadian rhythm dysregulation on health.
URL:https://www.ibs.re.kr/bimag/event/circadian-rhythm-profiles-derived-from-accelerometer-measures-of-the-sleep-wake-cycle-in-two-cohort-studies-chitaranjan-mahapatra/
LOCATION:B232 Seminar Room\, IBS\, 55 Expo-ro Yuseong-gu\, Daejeon\, Daejeon\, 34126\, Korea\, Republic of
CATEGORIES:Journal Club
ORGANIZER;CN="Jae Kyoung Kim":MAILTO:jaekkim@kaist.ac.kr
END:VEVENT
BEGIN:VEVENT
DTSTART;TZID=Asia/Seoul:20260605T100000
DTEND;TZID=Asia/Seoul:20260605T120000
DTSTAMP:20260527T003414Z
CREATED:20260429T070216Z
LAST-MODIFIED:20260527T003414Z
UID:12396-1780653600-1780660800@www.ibs.re.kr
SUMMARY:Chaos Meets Attention: Transformers for Large-Scale Dynamical Prediction- Gyuyoung Hwang
DESCRIPTION:In this talk\, we discuss the paper “Chaos Meets Attention: Transformers for Large-Scale Dynamical Prediction” by Yi He et al.\, ICML Poster\, 2025. \nAbstract: \nGenerating long-term trajectories of dissipative chaotic systems autoregressively is a highly challenging task. The inherent positive Lyapunov exponents amplify prediction errors over time. Many chaotic systems possess a crucial property — ergodicity on their attractors\, which makes long-term prediction possible. State-of-the-art methods address ergodicity by preserving statistical properties using optimal transport techniques. However\, these methods face scalability challenges due to the curse of dimensionality when matching distributions. To overcome this bottleneck\, we propose a scalable transformer-based framework capable of stably generating long-term high-dimensional and high-resolution chaotic dynamics while preserving ergodicity. Our method is grounded in a physical perspective\, revisiting the Von Neumann mean ergodic theorem to ensure the preservation of long-term statistics in the L2 space. We introduce novel modifications to the attention mechanism\, making the transformer architecture well-suited for learning large-scale chaotic systems. Compared to operator-based and transformer-based methods\, our model achieves better performances across five metrics\, from short-term prediction accuracy to long-term statistics. In addition to our methodological contributions\, we introduce new chaotic system benchmarks: a machine learning dataset of 140 snapshots of turbulent channel flow and a processed high-dimensional Kolmogorov Flow dataset\, along with various evaluation metrics for both short- and long-term performances. Both are well-suited for machine learning research on chaotic systems.
URL:https://www.ibs.re.kr/bimag/event/bridging-known-and-unknown-dynamics-by-transformer-based-machine-learning-inference-from-sparse-observations-gyuyoung-hwang/
LOCATION:B232 Seminar Room\, IBS\, 55 Expo-ro Yuseong-gu\, Daejeon\, Daejeon\, 34126\, Korea\, Republic of
CATEGORIES:Journal Club
ORGANIZER;CN="Jae Kyoung Kim":MAILTO:jaekkim@kaist.ac.kr
END:VEVENT
BEGIN:VEVENT
DTSTART;TZID=Asia/Seoul:20260619T100000
DTEND;TZID=Asia/Seoul:20260619T120000
DTSTAMP:20260614T040037Z
CREATED:20260520T075146Z
LAST-MODIFIED:20260614T040037Z
UID:12428-1781863200-1781870400@www.ibs.re.kr
SUMMARY:Inferring circadian phases and quantifying biological desynchrony across single-cell transcriptomes - Dongju Lim
DESCRIPTION:In this talk\, we discuss the paper “Inferring circadian phases and quantifying biological desynchrony across single-cell transcriptomes” by Andrea Salati et al.\, bioRxiv\, 2026. \n  \nAbstract: \nSingle-cell RNA sequencing (scRNA-seq) reveals heterogeneity in circadian clock states across individual cells\, yet accurately inferring circadian phase and distinguishing biological desynchrony from technical noise remains challenging. Here\, we introduce scRitmo\, a probabilistic framework that infers single-cell circadian phases from mRNA count data\, providing both a point estimate and a posterior uncertainty for each cell. A simulationcalibrated variance decomposition separates the observed phase dispersion into biological and technical components\, enabling direct estimation of intercellular desynchrony. We validate scRitmo using deeply sequenced unsynchronized fibroblasts\, where inferred transcriptomic phases accurately predict protein-level oscillations of a circadian reporter. Applied to murine scRNA-seq datasets from liver\, aorta\, and skin\, scRitmo outperforms existing methods and reveals cell-type-specific levels of phase coherence. In SABER-FISH time-series data\, the method recovers the progressive accumulation of desynchrony following synchronization\, and in Drosophila clock neurons it captures cell-type-specific phase shifts and the expected increase in phase dispersion under constant darkness relative to light-dark entrainment. Together\, scRitmo provides a principled approach for quantifying circadian (de)synchrony from transcriptomic data\, decoupling biological phase variability from measurement noise across tissues\, organisms\, and experimental conditions.
URL:https://www.ibs.re.kr/bimag/event/insulin-resistance-prediction-from-wearables-and-routine-blood-biomarkers-dongju-lim/
LOCATION:B232 Seminar Room\, IBS\, 55 Expo-ro Yuseong-gu\, Daejeon\, Daejeon\, 34126\, Korea\, Republic of
CATEGORIES:Journal Club
ORGANIZER;CN="Jae Kyoung Kim":MAILTO:jaekkim@kaist.ac.kr
END:VEVENT
BEGIN:VEVENT
DTSTART;TZID=Asia/Seoul:20260626T093000
DTEND;TZID=Asia/Seoul:20260626T113000
DTSTAMP:20260625T033224Z
CREATED:20260528T012227Z
LAST-MODIFIED:20260625T033224Z
UID:12546-1782466200-1782473400@www.ibs.re.kr
SUMMARY:Insulin resistance prediction from wearables and routine blood biomarkers - Hyunji Jeong
DESCRIPTION:In this talk\, we discuss the paper “Insulin resistance prediction from wearables and routine blood biomarkers” by Ahmed A. Metwally et al.\, Nature\, 2026. \nAbstract: \nInsulin resistance (IR)\, a primary precursor to type 2 diabetes\, is characterized by impaired insulin action in tissues1. However\, diagnostic methods remain expensive and inaccessible\, which hinders early intervention2\,3. Here we present the WEAR-ME study\, a large\, remotely conducted study of IR (n = 1\,165 participants; median body mass index (BMI) = 28 kg m−2\, median age = 45 years\, median haemoglobin A1c (HbA1c) = 5.4%) that uses time-series data from wearable devices and routine blood biomarkers to train deep neural networks against a ground-truth measure of IR (homeostatic model assessment of IR; HOMA-IR). Using a HOMA-IR cut-off of 2.9\, our multimodal model achieved robust performance (area under the receiver operating characteristic curve (AUROC) = 0.80\, sensitivity = 76%\, specificity = 84%) with data from wearable devices\, together with demographic and routine blood biomarker data. To enhance the use of time-series data from wearables\, we fine-tuned a wearable foundation model (WFM) pretrained on 40 million hours of sensor data. In an independent validation cohort (n = 72)\, a model integrating WFM-derived representations with demographic data surpassed a demographics-only baseline (AUROC = 0.75 versus 0.66). Moreover\, adding WFM-derived representations to a model with demographics\, fasting glucose and a lipid panel substantially improved performance\, compared with an identical model without data from wearables (AUROC = 0.88 versus 0.76). We integrate IR prediction into a large language model to contextualize the results and facilitate personalized recommendations. This work establishes a scalable\, accessible framework for the early detection of metabolic risk\, which could enable timely lifestyle interventions to prevent progression to type 2 diabetes.
URL:https://www.ibs.re.kr/bimag/event/learning-longitudinal-health-representations-from-ehr-and-wearable-data-hyunji-jeong/
LOCATION:B232 Seminar Room\, IBS\, 55 Expo-ro Yuseong-gu\, Daejeon\, Daejeon\, 34126\, Korea\, Republic of
CATEGORIES:Journal Club
ORGANIZER;CN="Jae Kyoung Kim":MAILTO:jaekkim@kaist.ac.kr
END:VEVENT
BEGIN:VEVENT
DTSTART;TZID=Asia/Seoul:20260629T110000
DTEND;TZID=Asia/Seoul:20260629T120000
DTSTAMP:20260615T110616Z
CREATED:20260615T110616Z
LAST-MODIFIED:20260615T110616Z
UID:12608-1782730800-1782734400@www.ibs.re.kr
SUMMARY:Prediction of mood state change based on repeated functional brain imaging and mathematical modeling in premenstrual syndrome - Dayoung Yoon
DESCRIPTION:Abstract: \nAccurately predicting mood fluctuations in mood disorders is critical for early intervention and personalized treatment. This study developed a neurophysiologically grounded mood prediction model by integrating behavioral modeling\, electroencephalography\, functional magnetic resonance imaging (fMRI)\, and physiological data from wearable devices in premenstrual syndrome (PMS). First\, applying the active inference framework to a risk-taking behavioral task revealed that PMS is characterized by a significant reduction in policy precision during decision-making during the luteal phase. Rather than a failure in learning trajectories\, this reduction reflects impulsivity at the behavioral execution stage and closely correlates with a diminished amplitude of the contingent negative variation (CNV)—an event-related potential indicating pre-decision neural preparation. Second\, neural features extracted by applying cortical surface-based geometric eigenmodes to fMRI data successfully differentiated mood states in PMS. We confirmed that these neural features can be accounted for by the control energy required to maintain eigenmodes based on structural connectivity. Furthermore\, to overcome the cost and accessibility constraints of fMRI\, we constructed an encoder model that approximates fMRI-based latent brain states and predicts mood using only four circadian rhythm markers continuously collected from wearable devices. Finally\, the significant correlation between policy precision and the reduction in centro-parietal CNV amplitude was also significantly explained by the control energy of the eigenmodes. In conclusion\, this study presents a real-time\, personalized mood monitoring framework that is firmly grounded in neurobiological mechanisms yet practically applicable to daily life.
URL:https://www.ibs.re.kr/bimag/event/prediction-of-mood-state-change-based-on-repeated-functional-brain-imaging-and-mathematical-modeling-in-premenstrual-syndrome-dayoung-yoon/
LOCATION:B232 Seminar Room\, IBS\, 55 Expo-ro Yuseong-gu\, Daejeon\, Daejeon\, 34126\, Korea\, Republic of
CATEGORIES:Biomedical Mathematics Seminar
ORGANIZER;CN="Jae Kyoung Kim":MAILTO:jaekkim@kaist.ac.kr
END:VEVENT
BEGIN:VEVENT
DTSTART;TZID=Asia/Seoul:20260703T100000
DTEND;TZID=Asia/Seoul:20260703T120000
DTSTAMP:20260528T012333Z
CREATED:20260527T140141Z
LAST-MODIFIED:20260528T012333Z
UID:12539-1783072800-1783080000@www.ibs.re.kr
SUMMARY:A Metabolism-Informed Neural Network Identifies Pathways Influencing the Potency and Toxicity of Antimicrobial Combinations - Se Jun Ahn
DESCRIPTION:In this talk\, we discuss the paper “A Metabolism-Informed Neural Network Identifies Pathways Influencing the Potency and Toxicity of Antimicrobial Combinations” by Harkirat Sigh Arora et al.\, npj drug discovery\, 2026. \nAbstract: \nAntimicrobial resistance poses a major global threat\, driven by diminishing efficacy of current treatments and limited new therapies. Combination therapy with existing drugs offers a promising solution\, yet current empirical screening methods are expensive and often lead to suboptimal efficacy and inadvertent toxicity. We introduce CALMA\, a computational framework that quantitatively analyzes the potency-toxicity landscape of multi-drug combinations. Integrating genome-scale metabolic modeling with a neural network that reflects metabolic subsystems\, CALMA enhances interpretability and prioritizes pathways influencing drug interactions. The incorporation of metabolic architecture in the neural network leads to over 92% reduction in model parameters\, enabling it to learn generalizable mechanistic signals and reducing the experimental search space of optimal combinations by 97%. CALMA identified promising antimicrobial combinations against Escherichia coli and Mycobacterium tuberculosis that were antagonistic for kidney and liver toxicity and uncovered the nucleotide salvage pathway as a selective influencer of toxicity\, which was validated in vitro. Mining of health records of over 400\,000 patients showed reduced frequency of kidney side-effects in patients taking a vancomycin combination identified by CALMA. CALMA provides a rational\, mechanistic approach to streamline combination treatment design.
URL:https://www.ibs.re.kr/bimag/event/a-metabolism-informed-neural-network-identifies-pathways-influencing-the-potency-and-toxicity-of-antimicrobial-combinations-se-jun-ahn/
LOCATION:B232 Seminar Room\, IBS\, 55 Expo-ro Yuseong-gu\, Daejeon\, Daejeon\, 34126\, Korea\, Republic of
CATEGORIES:Journal Club
ORGANIZER;CN="Jae Kyoung Kim":MAILTO:jaekkim@kaist.ac.kr
END:VEVENT
BEGIN:VEVENT
DTSTART;TZID=Asia/Seoul:20260706T100000
DTEND;TZID=Asia/Seoul:20260706T110000
DTSTAMP:20260701T084146Z
CREATED:20260616T012906Z
LAST-MODIFIED:20260701T084146Z
UID:12611-1783332000-1783335600@www.ibs.re.kr
SUMMARY:The effect of the fitness gradient - Jakub Svoboda
DESCRIPTION:Abstract: \nEvolutionary biology studies populations of reproducing individuals and how their composition changes over time.An important question is the fixation probability of a single mutant that attempts to invade a homogeneous population.Many real populations experience gradients of chemicals or nutrients that cause mutations to be beneficial in some spatial regions and harmful in others.We will examine the fixation probability of a mutant placed on a simple one-dimensional spatial structure that experiences such a gradient.The mutant’s fitness varies linearly but is on average 1\, whereas the resident’s fitness is constant and equal to 1.We will prove nonintuitive results about the fixation probability of mutants.
URL:https://www.ibs.re.kr/bimag/event/the-effect-of-the-fitness-gradient-jakub-svoboda/
LOCATION:B232 Seminar Room\, IBS\, 55 Expo-ro Yuseong-gu\, Daejeon\, Daejeon\, 34126\, Korea\, Republic of
CATEGORIES:Biomedical Mathematics Seminar
END:VEVENT
BEGIN:VEVENT
DTSTART;TZID=Asia/Seoul:20260709T100000
DTEND;TZID=Asia/Seoul:20260709T110000
DTSTAMP:20260708T001039Z
CREATED:20260708T001039Z
LAST-MODIFIED:20260708T001039Z
UID:12695-1783591200-1783594800@www.ibs.re.kr
SUMMARY:Advanced Iterative Methods as Elementary Iterations on Larger Spaces - Jongho Park
DESCRIPTION:Abstract: \nA central goal of scientific computing is to develop accurate and efficient solvers for scientific problems\, and this goal is often pursued through sophisticated numerical methods. In modern machine learning\, by contrast\, the basic optimization procedure is often comparatively simple\, typically gradient descent and its variants\, while much of the complexity is shifted to larger models. This talk examines this contrast from the viewpoint of scientific computing.We show that many advanced iterative methods\, including domain decomposition and multigrid methods\, can be interpreted as elementary iterations applied to equivalent problems posed on larger spaces. For example\, a classical multigrid method can be viewed as a Gauss–Seidel iteration for a suitable expanded system associated with a multilevel frame. To make this interpretation rigorous\, we introduce an auxiliary-space framework that recasts an iterative method for the original system as an equivalent\, but more elementary\, method for a lifted auxiliary system.The framework applies to a broad range of advanced methods. We illustrate its utility through applications to various modern iterative methods. Finally\, we discuss how this viewpoint can inform the design of numerical methods for problems arising in machine learning.
URL:https://www.ibs.re.kr/bimag/event/advanced-iterative-methods-as-elementary-iterations-on-larger-spaces-jongho-park/
LOCATION:B232 Seminar Room\, IBS\, 55 Expo-ro Yuseong-gu\, Daejeon\, Daejeon\, 34126\, Korea\, Republic of
CATEGORIES:Biomedical Mathematics Seminar
ORGANIZER;CN="Jae Kyoung Kim":MAILTO:jaekkim@kaist.ac.kr
END:VEVENT
BEGIN:VEVENT
DTSTART;TZID=Asia/Seoul:20260710T100000
DTEND;TZID=Asia/Seoul:20260710T120000
DTSTAMP:20260629T080759Z
CREATED:20260629T080759Z
LAST-MODIFIED:20260629T080759Z
UID:12659-1783677600-1783684800@www.ibs.re.kr
SUMMARY:Topological identification and interpretation for single-cell gene regulation elucidation across multiple platforms using scMGCA - Yun Min Song
DESCRIPTION:In this talk\, we discuss the paper “Topological identification and interpretation for single-cell gene regulation elucidation across multiple platforms using scMGCA” by Zhuohan Yu et al.\, nature communications\, 2023. \nAbstract: \n\n\n\nSingle-cell RNA sequencing provides high-throughput gene expression information to explore cellular heterogeneity at the individual cell level. A major challenge in characterizing high-throughput gene expression data arises from challenges related to dimensionality\, and the prevalence of dropout events. To address these concerns\, we develop a deep graph learning method\, scMGCA\, for single-cell data analysis. scMGCA is based on a graph-embedding autoencoder that simultaneously learns cell-cell topology representation and cluster assignments. We show that scMGCA is accurate and effective for cell segregation and batch effect correction\, outperforming other state-of-the-art models across multiple platforms. In addition\, we perform genomic interpretation on the key compressed transcriptomic space of the graph-embedding autoencoder to demonstrate the underlying gene regulation mechanism. We demonstrate that in a pancreatic ductal adenocarcinoma dataset\, scMGCA successfully provides annotations on the specific cell types and reveals differential gene expression levels across multiple tumor-associated and cell signalling pathways.
URL:https://www.ibs.re.kr/bimag/event/topological-identification-and-interpretation-for-single-cell-gene-regulation-elucidation-across-multiple-platforms-using-scmgca-yun-min-song/
LOCATION:B232 Seminar Room\, IBS\, 55 Expo-ro Yuseong-gu\, Daejeon\, Daejeon\, 34126\, Korea\, Republic of
CATEGORIES:Journal Club
ORGANIZER;CN="Jae Kyoung Kim":MAILTO:jaekkim@kaist.ac.kr
END:VEVENT
BEGIN:VEVENT
DTSTART;TZID=Asia/Seoul:20260724T100000
DTEND;TZID=Asia/Seoul:20260724T120000
DTSTAMP:20260713T062127Z
CREATED:20260713T062127Z
LAST-MODIFIED:20260713T062127Z
UID:12700-1784887200-1784894400@www.ibs.re.kr
SUMMARY:Topology identifies concurrent cyclic processes in single-cell transcriptomics and androgen receptor function - Seongjin Choi
DESCRIPTION:In this talk\, we discuss the paper “Topology identifies concurrent cyclic processes in single-cell transcriptomics and androgen receptor function” by Kelly Maggs et al.\, bioRxiv\, 2025. \nAbstract: \n\nStandard single-cell RNA-seq analysis frameworks aggregate over-lapping biological processes and impose a single parametrization\, conflating distinct programs. Here\, we introduce a topological framework that detects and disentangles multiple cyclic processes directly from single-cell transcriptomic data. We validate this approach on synthetic datasets and scRNA-seq profiles of human dermal fibroblasts under control conditions and following androgen receptor (AR) silencing\, as well as in vivo mouse prostate regeneration under androgen receptor add-back. We show robust cell cycle structure across conditions\, identify an unbiased AR-linked stress signature related to the senescence and proliferation across organisms\, and uncover cholesterol homeostasis as an AR-linked program in tissue regeneration. This framework enables identification and separation of concurrent cyclic processes from snapshot single-cell data\, revealing complex multi-dimensional regulatory dynamics inaccessible to standard clustering analysis.
URL:https://www.ibs.re.kr/bimag/event/topology-identifies-concurrent-cyclic-processes-in-single-cell-transcriptomics-and-androgen-receptor-function-seongjin-choi/
LOCATION:B232 Seminar Room\, IBS\, 55 Expo-ro Yuseong-gu\, Daejeon\, Daejeon\, 34126\, Korea\, Republic of
CATEGORIES:Journal Club
ORGANIZER;CN="Jae Kyoung Kim":MAILTO:jaekkim@kaist.ac.kr
END:VEVENT
BEGIN:VEVENT
DTSTART;TZID=Asia/Seoul:20260728T160000
DTEND;TZID=Asia/Seoul:20260728T170000
DTSTAMP:20260713T132333Z
CREATED:20260713T061641Z
LAST-MODIFIED:20260713T132333Z
UID:12698-1785254400-1785258000@www.ibs.re.kr
SUMMARY:Global Linearization of Nonlinear Dynamics via Koopman Operators: A Gentle Introduction\, Applications\, and Open Challenges - Hyukpyo Hong
DESCRIPTION:Abstract: \nA central challenge of modern dynamical systems theory is to make nonlinear systems tractable without sacrificing fidelity. Koopman operator theory pursues this goal by lifting nonlinear dynamics into a linear\, but infinite dimensional\, operator acting on a function space. This operator-theoretic perspective underlies a broad class of modern data-driven methods\, from dynamic mode decomposition to equation discovery in scientific machine learning for fluid dynamics and neuroscience. Yet this power comes at a price: the operator’s infinite dimensionality poses a fundamental obstacle to computation and practical use\, and finding tractable finite-dimensional approximations remains an open and active challenge. In this talk\, I will first introduce the basic principles of Koopman operator theory and survey some of the results that have made it a cornerstone of modern dynamical systems analysis. I will then briefly describe two of my works on finite-dimensional Koopman representations. Finally\, I will turn to my recent work on non-autonomous dynamical system learning\, in collaboration with Prof. Dae Wook Kim.
URL:https://www.ibs.re.kr/bimag/event/global-linearization-of-nonlinear-dynamics-via-koopman-operators-a-gentle-introduction-applications-and-open-challenges-hyukpyo-hong/
LOCATION:108\, Conference Room\, IBS\, 55 Expo-ro Yuseong-gu\, Daejeon\, Korea\, Republic of
CATEGORIES:Biomedical Mathematics Seminar
ORGANIZER;CN="Jae Kyoung Kim":MAILTO:jaekkim@kaist.ac.kr
END:VEVENT
BEGIN:VEVENT
DTSTART;TZID=Asia/Seoul:20260731T100000
DTEND;TZID=Asia/Seoul:20260731T120000
DTSTAMP:20260730T020723Z
CREATED:20260629T081120Z
LAST-MODIFIED:20260730T020723Z
UID:12661-1785492000-1785499200@www.ibs.re.kr
SUMMARY:Fast training of accurate physics-informed neural networks without gradient descent - Seunghun Lee
DESCRIPTION:In this talk\, we discuss the paper “Fast training of accurate physics-informed neural networks without gradient descent” by Chinmay Datar et al.\, ICLR\, 2026. \nAbstract: \nSolving time-dependent Partial Differential Equations (PDEs) is one of the most critical problems in computational science. While Physics-Informed Neural Networks (PINNs) offer a promising framework for approximating PDE solutions\, their accuracy and training speed are limited by two core barriers: gradient-descent-based iterative optimization over complex loss landscapes and non-causal treatment of time as an extra spatial dimension. We present Frozen-PINN\, a novel PINN based on the principle of space-time separation that leverages random features instead of training with gradient descent\, and incorporates temporal causality by construction. On nine PDE benchmarks\, including challenges like extreme advection speeds\, shocks\, and high-dimensionality\, Frozen-PINNs achieve superior training efficiency and accuracy over state-of-the-art PINNs\, often by several orders of magnitude. Our work addresses longstanding training and accuracy bottlenecks of PINNs\, delivering quickly trainable\, highly accurate\, and inherently causal PDE solvers\, a combination that prior methods could not realize. Our approach challenges the reliance of PINNs on stochastic gradient-descent-based methods and specialized hardware\, leading to a paradigm shift in PINN training and providing a challenging benchmark for the community.
URL:https://www.ibs.re.kr/bimag/event/oscillatory-network-efficiency-predicts-mood-and-fatigue-during-sleep-deprivation-seunghun-lee/
LOCATION:108\, Conference Room\, IBS\, 55 Expo-ro Yuseong-gu\, Daejeon\, Korea\, Republic of
CATEGORIES:Journal Club
ORGANIZER;CN="Jae Kyoung Kim":MAILTO:jaekkim@kaist.ac.kr
END:VEVENT
BEGIN:VEVENT
DTSTART;TZID=Asia/Seoul:20260803T110000
DTEND;TZID=Asia/Seoul:20260803T120000
DTSTAMP:20260728T040536Z
CREATED:20260728T040536Z
LAST-MODIFIED:20260728T040536Z
UID:12743-1785754800-1785758400@www.ibs.re.kr
SUMMARY:Infection dynamics at the host and cellular levels - Seong Jun Park
DESCRIPTION:Abstract: \nIn general\, the rates of infection and removal (whether through recovery or death) are nonlinear functions of the number of infected and susceptible individuals. One of the simplest models for the spread of infectious diseases is the SIR model\, which categorizes individuals as susceptible\, infectious\, recovered or deceased. In this model\, the infection rate\, governing the transition from susceptible to infected individuals\, is given by a linear function of both susceptible and infected populations. Similarly\, the removal rate\, representing the transition from infected to removed individuals\, is a linear function of the number of infected individuals. However\, existing research often overlooks the impact of nonlinear infection and removal rates in infection dynamics. This work presents an analytic expression for the number of infected individuals considering nonlinear infection and removal rates. In particular\, we examine how the number of infected individuals varies as cases emerge and obtain the expression accounting for the number of infected individuals at each moment. Viruses are microscopic infectious agents that require a host cell for replication. Viral replication occurs in several stages\, and the completion time for each stage varies due to differences in the cellular environment. Thus\, the time to complete each stage in viral replication is a random variable. However\, no analytic expression exists for the viral population at the cellular level when the completion time for each process constituting viral replication is a random variable. This study presents a simplified model of viral replication\, treating each stage as a renewal process with independently and identically distributed completion times. Using the proposed model\, we derive an analytical formula for viral populations at the cellular level\, based on viewing viral replication as a birth-death process. The mean viral count is expressed via probability density functions representing the completion time for each step in the replication process. This work validates the results with stochastic simulations. This study provides a new quantitative framework for understanding viral infection dynamics at host and cellular levels.
URL:https://www.ibs.re.kr/bimag/event/infection-dynamics-at-the-host-and-cellular-levels-seong-jun-park/
LOCATION:B232 Seminar Room\, IBS\, 55 Expo-ro Yuseong-gu\, Daejeon\, Daejeon\, 34126\, Korea\, Republic of
CATEGORIES:Biomedical Mathematics Seminar
ORGANIZER;CN="Jae Kyoung Kim":MAILTO:jaekkim@kaist.ac.kr
END:VEVENT
BEGIN:VEVENT
DTSTART;TZID=Asia/Seoul:20260807T100000
DTEND;TZID=Asia/Seoul:20260807T120000
DTSTAMP:20260730T084420Z
CREATED:20260730T084420Z
LAST-MODIFIED:20260730T084420Z
UID:12778-1786096800-1786104000@www.ibs.re.kr
SUMMARY:AutoML-Agent: A Multi-Agent LLM Framework for Full-Pipeline AutoML - Jin Woo Hyun
DESCRIPTION:In this talk\, we discuss the paper “AutoML-Agent: A Multi-Agent LLM Framework for Full-Pipeline AutoML” by Patara Trirat et al\, ICML\, 2025. \nAbstract: \nAutomated machine learning (AutoML) accelerates AI development by automating tasks in the development pipeline\, such as optimal model search and hyperparameter tuning. Existing AutoML systems often require technical expertise to set up complex tools\, which is in general time-consuming and requires a large amount of human effort. Therefore\, recent works have started exploiting large language models (LLM) to lessen such burden and increase the usability of AutoML frameworks via a natural language interface\, allowing non-expert users to build their data-driven solutions. These methods\, however\, are usually designed only for a particular process in the AI development pipeline and do not efficiently use the inherent capacity of the LLMs. This paper proposes AutoML-Agent\, a novel multi-agent framework tailored for full-pipeline AutoML\, i.e.\, from data retrieval to model deployment. AutoML-Agent takes user’s task descriptions\, facilitates collaboration between specialized LLM agents\, and delivers deployment-ready models. Unlike existing work\, instead of devising a single plan\, we introduce a retrieval-augmented planning strategy to enhance exploration to search for more optimal plans. We also decompose each plan into sub-tasks (e.g.\, data preprocessing and neural network design) each of which is solved by a specialized agent we build via prompting executing in parallel\, making the search process more efficient. Moreover\, we propose a multi-stage verification to verify executed results and guide the code generation LLM in implementing successful solutions. Extensive experiments on seven downstream tasks using fourteen datasets show that AutoML-Agent achieves a higher success rate in automating the full AutoML process\, yielding systems with good performance throughout the diverse domains.
URL:https://www.ibs.re.kr/bimag/event/automl-agent-a-multi-agent-llm-framework-for-full-pipeline-automl-jin-woo-hyun/
LOCATION:B232 Seminar Room\, IBS\, 55 Expo-ro Yuseong-gu\, Daejeon\, Daejeon\, 34126\, Korea\, Republic of
CATEGORIES:Journal Club
ORGANIZER;CN="Jae Kyoung Kim":MAILTO:jaekkim@kaist.ac.kr
END:VEVENT
BEGIN:VEVENT
DTSTART;TZID=Asia/Seoul:20260814T100000
DTEND;TZID=Asia/Seoul:20260814T120000
DTSTAMP:20260730T035435Z
CREATED:20260730T035435Z
LAST-MODIFIED:20260730T035435Z
UID:12774-1786701600-1786708800@www.ibs.re.kr
SUMMARY:Evolution of error correction through a need for speed - Kangmin Lee
DESCRIPTION:In this talk\, we discuss the paper “Evolution of error correction through a need for speed” by Riccardo Ravasio et al.\, Science\, 2026. \nAbstract: \nKinetic proofreading is a class of error-correcting mechanisms in biology that expend energy to avoid mistakes during replication\, transcription\, and translation. Proofreading is typically assumed to evolve when selection for fidelity outweighs costs in energy and the speed of replication. We show that when stalling after misincorporations is accounted for\, proofreading can instead speed up replication. Consistent with data on polymerase mutagenesis\, our results suggest that proofreading can evolve under selection for speed alone. We generalize to multicomponent self-assembly and show that analogous error-correcting processes\, such as dynamic instability\, can likewise emerge purely from selection for rapid assembly. Thus\, nonequilibrium error correction can evolve from selection for speed\, even without direct fidelity advantages. We discuss implications for mutation-rate evolution\, molecular assembly processes\, and models of early life.
URL:https://www.ibs.re.kr/bimag/event/evolution-of-error-correction-through-a-need-for-speed-kangmin-lee/
LOCATION:108\, Conference Room\, IBS\, 55 Expo-ro Yuseong-gu\, Daejeon\, Korea\, Republic of
CATEGORIES:Journal Club
ORGANIZER;CN="Jae Kyoung Kim":MAILTO:jaekkim@kaist.ac.kr
END:VEVENT
BEGIN:VEVENT
DTSTART;TZID=Asia/Seoul:20260817T160000
DTEND;TZID=Asia/Seoul:20260817T170000
DTSTAMP:20260810T020452Z
CREATED:20260810T020342Z
LAST-MODIFIED:20260810T020452Z
UID:12808-1786982400-1786986000@www.ibs.re.kr
SUMMARY:수학이 공학의 언어라면 - 민태기
DESCRIPTION:  \n이번 세미나에서는 “조선이 만난 아인슈타인”\, “판타 레이” 등을 저술하신 민태기 소장님을 모시고 “수학이 공학의 언어라면”이라는 주제로 강연을 진행할 예정입니다.
URL:https://www.ibs.re.kr/bimag/event/%ec%88%98%ed%95%99%ec%9d%b4-%ea%b3%b5%ed%95%99%ec%9d%98-%ec%96%b8%ec%96%b4%eb%9d%bc%eb%a9%b4-%eb%af%bc%ed%83%9c%ea%b8%b0/
LOCATION:109\, Conference room\, IBS\, 55 Expo-ro Yuseong-gu\, Daejeon\, Korea\, Republic of
CATEGORIES:Biomedical Mathematics Seminar
ATTACH;FMTTYPE=image/jpeg:https://www.ibs.re.kr/bimag/cms/wp-content/uploads/2026/08/L20230605095029.jpg
ORGANIZER;CN="Jae Kyoung Kim":MAILTO:jaekkim@kaist.ac.kr
END:VEVENT
BEGIN:VEVENT
DTSTART;TZID=Asia/Seoul:20260821T100000
DTEND;TZID=Asia/Seoul:20260821T120000
DTSTAMP:20260730T035554Z
CREATED:20260730T020529Z
LAST-MODIFIED:20260730T035554Z
UID:12772-1787306400-1787313600@www.ibs.re.kr
SUMMARY:Tangent-Space Regularization for Neural-Network Models of Dynamical Systems - Olive Cawiding
DESCRIPTION:In this talk\, we discuss the paper “Tangent-Space Regularization for Neural-Network Models of Dynamical Systems” by Fredrik B Carlson et al\, arXiv (2026). \nAbstract: \nThis work introduces the concept of tangent space regularization for neural-network models of dynamical systems. The tangent space to the dynamics function of many physical systems of interest in control applications exhibits useful properties\, e.g.\, smoothness\, motivating regularization of the model Jacobian along system trajectories using assumptions on the tangent space of the dynamics. Without assumptions\, large amounts of training data are required for a neural network to learn the full non-linear dynamics without overfitting. We compare different network architectures on one-step prediction and simulation performance and investigate the propensity of different architectures to learn models with correct input-output Jacobian. Furthermore\, the influence of L2 weight regularization on the learned Jacobian eigenvalue spectrum\, and hence system stability\, is investigated.
URL:https://www.ibs.re.kr/bimag/event/tangent-space-regularization-for-neural-network-models-of-dynamical-systems-olive-cawiding/
LOCATION:108\, Conference Room\, IBS\, 55 Expo-ro Yuseong-gu\, Daejeon\, Korea\, Republic of
CATEGORIES:Journal Club
ORGANIZER;CN="Jae Kyoung Kim":MAILTO:jaekkim@kaist.ac.kr
END:VEVENT
BEGIN:VEVENT
DTSTART;TZID=Asia/Seoul:20260828T100000
DTEND;TZID=Asia/Seoul:20260828T120000
DTSTAMP:20260730T040743Z
CREATED:20260730T040743Z
LAST-MODIFIED:20260730T040743Z
UID:12776-1787911200-1787918400@www.ibs.re.kr
SUMMARY:Towards a General Intelligence and Interface for Wearable Health Data - Aqsa Awan
DESCRIPTION:In this talk\, we discuss the paper “Towards a General Intelligence and Interface for Wearable Health Data” by Girish Narayanswamy et al.\, arXiv\, 2026. \nAbstract: \nWhile ubiquitous wearable sensors capture a wealth of behavioral and physiological information\, effectively transforming these signals into personalized health insights is challenging. Specifically\, converting low-level sensor data into representations capable of characterizing higher-level states is difficult due to high phenotypic diversity and variation in individual baseline health\, physiology\, and lifestyle factors. Moreover\, collecting wearable data paired with health outcome annotations is laborious and expensive\, and retrospective annotation remains practically unfeasible\, contributing to a scarcity of data with high-quality labels. To overcome these limitations\, we propose a foundation model for wearable health that is pretrained on more than one trillion minutes of unlabeled sensor signals drawn from a large cohort of five million participants. We demonstrate that the joint scaling of model capacity and pretraining data volume leads to systematic improvements in performance\, as evaluated on a diverse set of 35 health prediction tasks\, spanning cardiovascular\, metabolic\, sleep\, and mental health\, as well as lifestyle choices and demographic factors. We find that this population scale representation unlocks label-efficient few-shot learning and generative capabilities for robust daily metric estimation. To further leverage this learned representation\, we deploy a classroom of LLM agents to autonomously search the space of downstream predictive heads built on the model embeddings\, showing broad performance improvements that increase with LLM model capacity. Finally\, we show how integrating these downstream predictors into a Personal Health Agent can support model responses that are more relevant\, contextually aware\, and safe\, and we validate this via 1\,860 ratings from a cohort of clinicians.
URL:https://www.ibs.re.kr/bimag/event/towards-a-general-intelligence-and-interface-for-wearable-health-data-aqsa-awan/
LOCATION:108\, Conference Room\, IBS\, 55 Expo-ro Yuseong-gu\, Daejeon\, Korea\, Republic of
CATEGORIES:Journal Club
ORGANIZER;CN="Jae Kyoung Kim":MAILTO:jaekkim@kaist.ac.kr
END:VEVENT
END:VCALENDAR