Pacific Symposium on Biocomputing. Pacific Symposium on Biocomputing最新文献_第5页

Provenance Tracing in Network Diffusion Algorithms. 网络扩散算法中的溯源。

Pacific Symposium on Biocomputing. Pacific Symposium on Biocomputing Pub Date : 2026-01-01 DOI: 10.1142/9789819824755_0055

Nure Tasnina, Mark Crovella, Simon Kasif, T M Murali

{"title":"Provenance Tracing in Network Diffusion Algorithms.","authors":"Nure Tasnina, Mark Crovella, Simon Kasif, T M Murali","doi":"10.1142/9789819824755_0055","DOIUrl":"10.1142/9789819824755_0055","url":null,"abstract":"We propose a novel strategy for provenance tracing in random walk-based network diffusion algorithms, a problem that has been surprisingly overlooked in spite of the widespread use of diffusion algorithms in biological applications. Our path-based approach enables ranking paths by the magnitude of their contribution to each node's score, offering insight into how information propagates through a network. Building on this capability, we introduce two quantitative measures: (i) path-based effective diffusion, which evaluates how well a diffusion algorithm leverages the full topology of a network, and (ii) diffusion betweenness, which quantifies a node's importance in propagating scores. We applied our framework to SARSCoV- 2 protein interactors and human PPI networks. Provenance tracing of the Regularized Laplacian and Random Walk with Restart algorithms revealed that a substantial amount of a node's score is contributed via multi-edge paths, demonstrating that diffusion algorithms exploit the non-local structure of the network. Analysis of diffusion betweenness identified proteins playing a critical role in score propagation; proteins with high diffusion betweenness are enriched with essential human genes and interactors of other viruses, supporting the biological interpretability of the metric. Finally, in a signaling network composed of causal interactions between human proteins, the top contributing paths showed strong overlap with COVID-19-related pathways. These results suggest that our path-based framework offers valuable insight into diffusion algorithms and can serve as a powerful tool for interpreting diffusion scores in a biologically meaningful context, complementing existing module- or node-centric approaches in systems biology. The code is publicly available at https://github.com/n-tasnina/provenance-tracing.git under the GNU General Public License v3.0.","PeriodicalId":34954,"journal":{"name":"Pacific Symposium on Biocomputing. Pacific Symposium on Biocomputing","volume":"31 ","pages":"755-769"},"PeriodicalIF":0.0,"publicationDate":"2026-01-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"147310816","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

引用次数: 0

Workshop Introduction: Advances of AI Methods in Single Cell Spatial Omics. 工作坊简介：人工智能方法在单细胞空间组学中的进展。

Pacific Symposium on Biocomputing. Pacific Symposium on Biocomputing Pub Date : 2026-01-01 DOI: 10.1142/9789819824755_0061

Lana Garmire, Xiuwei Zhang, Joshua Levy

引用次数: 0

REPEL - Random Embedding Perturbation for Enhanced Learning of Protein Function. 用于增强蛋白质功能学习的随机嵌入微扰。

Pacific Symposium on Biocomputing. Pacific Symposium on Biocomputing Pub Date : 2026-01-01 DOI: 10.1142/9789819824755_0056

Di Zhou, Lenore J Cowen, Kaiyi Wu, Xiaozhe Hu, Donna K Slonim

{"title":"REPEL - Random Embedding Perturbation for Enhanced Learning of Protein Function.","authors":"Di Zhou, Lenore J Cowen, Kaiyi Wu, Xiaozhe Hu, Donna K Slonim","doi":"10.1142/9789819824755_0056","DOIUrl":"10.1142/9789819824755_0056","url":null,"abstract":"Protein function prediction from multiplex protein-protein association networks is a crucial approach to extending functional annotation. Current methods use embeddings of the heterogeneous network data that aim to place related proteins near each other in embedding space. However, such embeddings suffer from spurious protein proximity as well, reducing function prediction accuracy. Because heterogeneous input networks often have very different structures, it is hard to confidently declare proteins to be dissimilar using the network structure or the resulting embeddings. Here we address this problem with REPEL, a function prediction tool using a random graph augmentation method that applies a uniform weak force to push nodes apart. We assess this method on simulated networks with planted overlapping communities, as well as on real multiplex yeast and E.coli protein association networks. Surprisingly, we find that this method consistently improves protein function prediction over competing methods Mashup, deepNF, and BIONIC. The random repelling nature of the augmented graphs has a denoising effect on the learning process, distancing node pairs with spurious proximity while preserving true functional connections, thus increasing robustness. This graph augmentation principle may generalize to denoising and improving robustness in other graph-based learning algorithms.","PeriodicalId":34954,"journal":{"name":"Pacific Symposium on Biocomputing. Pacific Symposium on Biocomputing","volume":"31 ","pages":"770-783"},"PeriodicalIF":0.0,"publicationDate":"2026-01-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"https://www.ncbi.nlm.nih.gov/pmc/articles/PMC12952673/pdf/","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"147310957","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"OA","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

引用次数: 0

Evaluation of Large Language Models as Emergency Department Revisit Predictors. 大型语言模型作为急诊科重访预测因子的评估。

Pacific Symposium on Biocomputing. Pacific Symposium on Biocomputing Pub Date : 2026-01-01 DOI: 10.1142/9789819824755_0010

Emma Chen, Luyang Luo, Fatma Gunturkun, Sraavya Sambara, Rushil Arora, Boyang Tom Jin, Pranav Rajpurkar, David A Kim

{"title":"Evaluation of Large Language Models as Emergency Department Revisit Predictors.","authors":"Emma Chen, Luyang Luo, Fatma Gunturkun, Sraavya Sambara, Rushil Arora, Boyang Tom Jin, Pranav Rajpurkar, David A Kim","doi":"10.1142/9789819824755_0010","DOIUrl":"10.1142/9789819824755_0010","url":null,"abstract":"Large Language Models (LLMs) have shown promise in clinical reasoning and question answering, yet their effectiveness for real-world clinical prediction remains an open question. We present the first large-scale study evaluating LLMs for predicting 30-day emergency department (ED) revisits using 138,010 visits from the Adult Emergency Department at Stanford. We assessed two modeling paradigms: (1) direct prediction, where the LLM generates revisit risk assessments in natural language, and (2) embedding-based approaches that leverage LLM-derived vector representations (LLM2Vec) of patient data for downstream modeling. Retrieval augmentation improved direct prediction performance (e.g., Claude 3.7 F1 from 0.3755 (95% CI [0.3647, 0.3864]) to 0.4160 (95% CI [0.4024, 0.4294])), and embedding-based methods consistently outperformed direct approaches, with LLM2Vec achieving F1=0.4505 (95% CI [0.4345, 0.4666]). Despite having access to comprehensive structured and unstructured clinical data, all LLM approaches (F1=0.3022-0.4505) failed to exceed a traditional LightGBM model using only structured data (F1=0.4614 (95% CI [0.4496, 0.4789])). Through systematic analysis of the reasoning chains in 17,488 predictions, we suggest potential failure patterns: reasoning may systematically degrade performance through overweighting medical histories and similar visits, neglecting protective factors, and risk aversion. Our work establishes essential baseline performance while revealing fundamental limitations in current-generation LLMs for clinical prediction tasks.","PeriodicalId":34954,"journal":{"name":"Pacific Symposium on Biocomputing. Pacific Symposium on Biocomputing","volume":"31 ","pages":"130-143"},"PeriodicalIF":0.0,"publicationDate":"2026-01-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"https://www.ncbi.nlm.nih.gov/pmc/articles/PMC12952678/pdf/","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"147310681","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"OA","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

引用次数: 0

AI for Health: Leveraging Artificial Intelligence to Revolutionize Healthcare. 人工智能健康：利用人工智能彻底改变医疗保健。

Pacific Symposium on Biocomputing. Pacific Symposium on Biocomputing Pub Date : 2026-01-01 DOI: 10.1142/9789819824755_0062

Ruowang Li, Brian D Davison, Tiffani Bright, Lifang He

引用次数: 0

The Evolving Cyberinfrastructure at the National Institutes of Health to Support Data and AI in Biomedical Research. 国家卫生研究院不断发展的网络基础设施，以支持生物医学研究中的数据和人工智能。

Pacific Symposium on Biocomputing. Pacific Symposium on Biocomputing Pub Date : 2026-01-01 DOI: 10.1142/9789819824755_0064

Ojas A Ramwala, Nick Weber, Sean D Mooney

{"title":"The Evolving Cyberinfrastructure at the National Institutes of Health to Support Data and AI in Biomedical Research.","authors":"Ojas A Ramwala, Nick Weber, Sean D Mooney","doi":"10.1142/9789819824755_0064","DOIUrl":"10.1142/9789819824755_0064","url":null,"abstract":"Technological advancements have made biomedicine rich in data. With the generation of enormous volumes of biomedical and clinical data, it has become imperative to support biomedical computing investigators to utilize this wealth of biologically meaningful information. Moreover, advancements in Artificial Intelligence (AI) techniques, in conjunction with improved capabilities in implementing large-scale data processing pipelines, have led to the development of robust computational techniques and algorithms to solve complex biological problems. However, there are many challenges associated with providing researchers with secure systems for accessing biomedical data and computational resources that must be addressed. Establishing and maintaining an impactful data and AI ecosystem to support efforts in advancing biomedical research requires effective, scalable, and standardized information technology solutions, funding programs, and technical guidance that facilitate researchers in utilizing the state-of-the-art. The U.S. National Institutes of Health (NIH) has established novel initiatives to implement a cyberinfrastructure that democratizes secure access to large biomedical datasets and cloud-based computing resources, equipping biocomputing scientists to pursue pioneering research. This workshop will highlight the major issues restraining researchers' access to biomedical datasets and computing infrastructures and will cover the key components of the NIH's cyberinfrastructure aimed at advancing data science and AI research for biomedical applications.","PeriodicalId":34954,"journal":{"name":"Pacific Symposium on Biocomputing. Pacific Symposium on Biocomputing","volume":"31 ","pages":"859-864"},"PeriodicalIF":0.0,"publicationDate":"2026-01-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"147310729","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

引用次数: 0

Retrieval-Augmented Guardrails for AI-Drafted Patient-Portal Messages: Error Taxonomy Construction and Large-Scale Evaluation. 人工智能起草的患者门户信息的检索增强护栏：错误分类构建和大规模评估。

Pacific Symposium on Biocomputing. Pacific Symposium on Biocomputing Pub Date : 2026-01-01 DOI: 10.1142/9789819824755_0014

Wenyuan Chen, Fateme Nateghi Haredasht, Kameron C Black, François Grolleau, Emily Alsentzer, Jonathan H Chen, Stephen P Ma

{"title":"Retrieval-Augmented Guardrails for AI-Drafted Patient-Portal Messages: Error Taxonomy Construction and Large-Scale Evaluation.","authors":"Wenyuan Chen, Fateme Nateghi Haredasht, Kameron C Black, François Grolleau, Emily Alsentzer, Jonathan H Chen, Stephen P Ma","doi":"10.1142/9789819824755_0014","DOIUrl":"10.1142/9789819824755_0014","url":null,"abstract":"Asynchronous patient-clinician messaging via EHR portals is a growing source of clinician workload, prompting interest in large language models (LLMs) to assist with draft responses. However, LLM outputs may contain clinical inaccuracies, omissions, or tone mismatches, making robust evaluation essential. Our contributions are threefold: (1) we introduce a clinically grounded error ontology comprising 5 domains and 59 granular error codes, developed through inductive coding and expert adjudication; (2) we develop a Retrieval-Augmented Error Checking (RAEC) pipeline that leverages semantically similar historical message-response pairs to improve judgment quality; and (3) we provide a two-stage prompting architecture using DSPy to enable scalable, interpretable, and hierarchical error detection. Our approach assesses the quality of drafts both in isolation and with reference to similar past message-response pairs retrieved from institutional archives. Using a two-stage DSPy pipeline, we compared baseline and reference-enhanced evaluations on over 1,500 patient messages. Retrieval context improved error identification in domains such as clinical completeness and workflow appropriateness. Human validation on 100 messages demonstrated superior agreement (concordance = 50% vs. 33%) and performance (F1 = 0.500 vs. 0.256) of context-enhanced labels vs. baseline, supporting the use of our RAEC pipeline as AI guardrails for patient messaging.","PeriodicalId":34954,"journal":{"name":"Pacific Symposium on Biocomputing. Pacific Symposium on Biocomputing","volume":"31 ","pages":"189-204"},"PeriodicalIF":0.0,"publicationDate":"2026-01-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"147310885","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

引用次数: 0

Network optimal retrieval of sparse perturbations for steady-state control. 稳态控制中稀疏扰动的网络最优检索。

Pacific Symposium on Biocomputing. Pacific Symposium on Biocomputing Pub Date : 2026-01-01 DOI: 10.1142/9789819824755_0057

Krithika Krishnan, Tiange Shi, Satyam Kumar, Han Yu, Rachael Hageman Blair

{"title":"Network optimal retrieval of sparse perturbations for steady-state control.","authors":"Krithika Krishnan, Tiange Shi, Satyam Kumar, Han Yu, Rachael Hageman Blair","doi":"10.1142/9789819824755_0057","DOIUrl":"10.1142/9789819824755_0057","url":null,"abstract":"Prioritizing targeted perturbation experiments remains a central challenge in systems biology, where experimental constraints limit network manipulation. We introduce NORSP (Network Optimal Retrieval of Sparse Perturbations). This novel computational framework integrates network propagation with supervised subset selection to identify minimal perturbation sets that can shift a system from its initial to a desired steady state. NORSP leverages a sensitivity matrix derived solely from network topology, enabling control prediction without requiring full knowledge of system dynamics. Applicable to undirected, directed, and signed networks, NORSP accommodates a broad range of biological models and experimental scenarios. We validate its effectiveness using YBX1 knockdown transcriptomics data and 61 curated metabolic networks from the BioModels repository, demonstrating NORSP's robustness, scalability, and experimental relevance. Even under constraints that obscure true perturbations, the algorithm reliably infers alternative targets that achieve comparable control. Control is confirmed both in graphical approximations and through full dynamical model simulations. Overall, NORSP provides a practical and generalizable solution for steady-state control in complex biological systems, laying the foundation for multi-omics hypothesis generation and systems-level experimental design.","PeriodicalId":34954,"journal":{"name":"Pacific Symposium on Biocomputing. Pacific Symposium on Biocomputing","volume":"31 ","pages":"784-798"},"PeriodicalIF":0.0,"publicationDate":"2026-01-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"147310812","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

引用次数: 0

Deep Learning-based Classification of Patients with Postural Orthostatic Tachycardia Syndrome using Wearable ECG and Accelerometer Data. 基于深度学习的基于可穿戴ECG和加速度计数据的体位性心动过速综合征患者分类。

Pacific Symposium on Biocomputing. Pacific Symposium on Biocomputing Pub Date : 2026-01-01 DOI: 10.1142/9789819824755_0050

Hyunjun Choi, Nicholas Matsumoto, Xi Li, Debbie Teodorescu, Anxhela Kote, Min-Jing Yang, Xiao Liu, Miguel E Hernandez, Jason H Moore, Graciela Gonzalez Hernandez, Peng-Sheng Chen

{"title":"Deep Learning-based Classification of Patients with Postural Orthostatic Tachycardia Syndrome using Wearable ECG and Accelerometer Data.","authors":"Hyunjun Choi, Nicholas Matsumoto, Xi Li, Debbie Teodorescu, Anxhela Kote, Min-Jing Yang, Xiao Liu, Miguel E Hernandez, Jason H Moore, Graciela Gonzalez Hernandez, Peng-Sheng Chen","doi":"10.1142/9789819824755_0050","DOIUrl":"10.1142/9789819824755_0050","url":null,"abstract":"Postural Orthostatic Tachycardia Syndrome (POTS) is a chronic autonomic disorder characterized by chronic (> 3 months) orthostatic intolerance and an increase in heart rate (HR) of ≥ 30 beats per minute (bpm) without orthostatic hypotension. Traditional diagnostic approaches, such as the active standing or tilt-table test, are typically conducted under controlled clinical conditions, limiting their ability to capture the natural variability of symptoms and the intricate physiological responses occurring in daily life. These tests may cause patient discomfort, dizziness, nausea, or syncope. Furthermore, they are timeconsuming and cannot be used as a screening tool for POTS. To address these limitations, this study explored wearable devices that continuously collect physiological data-specifically, electrocardiogram (ECG) and accelerometer (ACC)-derived metrics-from POTS patients and healthy controls during routine daily activities. Physiological features around posturechange events identified in the data were processed and used to train and test a baseline deep learning model. The model demonstrated promising performance in accurately differentiating POTS patients from healthy controls in a relatively small cohort (66 from POTS patients and 20 from controls), indicating its potential as a feasibility study for clinical decision support. Future studies involving larger and more diverse samples under varying clinical conditions would be necessary to enhance the robustness and viability of our diagnostic model.","PeriodicalId":34954,"journal":{"name":"Pacific Symposium on Biocomputing. Pacific Symposium on Biocomputing","volume":"31 ","pages":"694-707"},"PeriodicalIF":0.0,"publicationDate":"2026-01-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"147310743","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

引用次数: 0

Session Introduction: AI and Machine Learning in Clinical Medicine Bridging or Separating Model Intelligence and Human Expertise. 会议介绍：临床医学中的人工智能和机器学习连接或分离模型智能和人类专业知识。

Pacific Symposium on Biocomputing. Pacific Symposium on Biocomputing Pub Date : 2026-01-01 DOI: 10.1142/9789819824755_0001

Fateme Nateghi Haredasht, Joseph D Romano, Brett K Beaulieu-Jones, Dokyoon Kim, Alexander Ioannidis, Geoffrey H Tison, Roxana Daneshjou, Jonathan H Chen

引用次数: 0