MOTIVATION: Genome-wide association studies (GWAS) at biobank scale are computationally intensive, especially for admixed populations requiring robust statistical models. SAIGE is a widely used method for generalized linear mixed-model GWAS but is limited by its CPU-based implementation, making phenome-wide association studies impractical for many research groups. RESULTS: We developed SAIGE-GPU, a GPU-accelerated version of SAIGE that replaces CPU-intensive matrix operations with GPU-optimized kernels. The core innovation is distributing genetic relationship matrix calculations across GPUs and communication layers. Applied to 2068 phenotypes from 635 969 participants in the Million Veteran Program, including diverse and admixed populations, SAIGE-GPU achieved a 5-fold speedup in mixed model fitting on supercomputing infrastructure and cloud platforms. We further optimized the variant association testing step through multi-core and multi-trait parallelization. Deployed on Google Cloud Platform and Azure, the method provided substantial cost and time savings. AVAILABILITY AND IMPLEMENTATION: Source code and binaries are available for download at https://github.com/saigegit/SAIGE/tree/SAIGE-GPU-1.3.3. A code snapshot is archived at Zenodo for reproducibility (DOI: [10.5281/zenodo.17642591]). SAIGE-GPU is available in a containerized format for use across HPC and cloud environments and is implemented in R/C++ and runs on Linux systems.

MOTIVATION: The increasing availability of Electronic Health Record (EHR) systems has created enormous potential for translational research. Recent developments in representation learning techniques have led to effective large-scale representations of EHR concepts along with knowledge graphs that empower downstream EHR studies. However, most existing methods require training with patient-level data, limiting their abilities to expand the training with multi-institutional EHR data. On the other hand, scalable approaches that only require summary-level data do not incorporate temporal dependencies between concepts. METHODS: We introduce a DirectiOnal Medical Embedding (DOME) algorithm to encode temporally directional relationships between medical concepts, using summary-level EHR data. Specifically, DOME first aggregates patient-level EHR data into an asymmetric co-occurrence matrix. Then it computes two Positive Pointwise Mutual Information (PPMI) matrices to correspondingly encode the pairwise prior and posterior dependencies between medical concepts. Following that, a joint matrix factorization is performed on the two PPMI matrices, which results in three vectors for each concept: a semantic embedding and two directional context embeddings. They collectively provide a comprehensive depiction of the temporal relationship between EHR concepts. RESULTS: We highlight the advantages and translational potential of DOME through three sets of validation studies. First, DOME consistently improves existing direction-agnostic embedding vectors for disease risk prediction in several diseases, for example achieving a relative gain of 5.5% in the area under the receiver operating characteristic (AUROC) for lung cancer. Second, DOME excels in directional drug-disease relationship inference by successfully differentiating between drug side effects and indications, correspondingly achieving relative AUROC gain over the state-of-the-art methods by 10.8% and 6.6%. Finally, DOME effectively constructs directional knowledge graphs, which distinguish disease risk factors from comorbidities, thereby revealing disease progression trajectories. The source codes are provided at https://github.com/celehs/Directional-EHR-embedding.

MOTIVATION: Predicting molecule-disease indications and side effects is important for drug development and pharmacovigilance. Comprehensively mining molecule-molecule, molecule-disease and disease-disease semantic dependencies can potentially improve prediction performance. METHODS: We introduce a Multi-Modal REpresentation Mapping Approach to Predicting molecular-disease relations (M2REMAP) by incorporating clinical semantics learned from electronic health records (EHR) of 12.6 million patients. Specifically, M2REMAP first learns a multimodal molecule representation that synthesizes chemical property and clinical semantic information by mapping molecule chemicals via a deep neural network onto the clinical semantic embedding space shared by drugs, diseases and other common clinical concepts. To infer molecule-disease relations, M2REMAP combines multimodal molecule representation and disease semantic embedding to jointly infer indications and side effects. RESULTS: We extensively evaluate M2REMAP on molecule indications, side effects and interactions. Results show that incorporating EHR embeddings improves performance significantly, for example, attaining an improvement over the baseline models by 23.6% in PRC-AUC on indications and 23.9% on side effects. Further, M2REMAP overcomes the limitation of existing methods and effectively predicts drugs for novel diseases and emerging pathogens. AVAILABILITY AND IMPLEMENTATION: The code is available at https://github.com/celehs/M2REMAP, and prediction results are provided at https://shiny.parse-health.org/drugs-diseases-dev/. SUPPLEMENTARY INFORMATION: Supplementary data are available at Bioinformatics online.

This tutorial focuses on both theoretical and practical aspects of improving the efficiency and robustness of transformer-based approaches, so that these can be effectively used in practical, high-scale, and high-volume information retrieval (IR) scenarios. The tutorial is inspired and informed by our work and experience while working with massive narrative datasets (8.5 billion medical notes), and by our basic research and academic experience with transformer-based IR tasks. Additionally, the tutorial focuses on techniques for making transformer-based IR robust against adversarial (AI) exploitation. This is a recent concern in the IR domain that we needed to take into concern, and we want to want to share some of the lessons learned and applicable principles with our audience. Finally, an important, if not critical, element of this tutorial is its focus on didacticism -- delivering tutorial content in a clear, intuitive, plain-speak fashion. Transformers are a challenging subject, and, through our teaching experience, we observed a great value and a great need to explain all relevant aspects of this architecture and related principles in the most straightforward, precise, and intuitive manner. That is the defining style of our proposed tutorial.

MOTIVATION: Biomedical machine reading comprehension (biomedical-MRC) aims to comprehend complex biomedical narratives and assist healthcare professionals in retrieving information from them. The high performance of modern neural network-based MRC systems depends on high-quality, large-scale, human-annotated training datasets. In the biomedical domain, a crucial challenge in creating such datasets is the requirement for domain knowledge, inducing the scarcity of labeled data and the need for transfer learning from the labeled general-purpose (source) domain to the biomedical (target) domain. However, there is a discrepancy in marginal distributions between the general-purpose and biomedical domains due to the variances in topics. Therefore, direct-transferring of learned representations from a model trained on a general-purpose domain to the biomedical domain can hurt the model's performance. RESULTS: We present an adversarial learning-based domain adaptation framework for the biomedical machine reading comprehension task (BioADAPT-MRC), a neural network-based method to address the discrepancies in the marginal distributions between the general and biomedical domain datasets. BioADAPT-MRC relaxes the need for generating pseudo labels for training a well-performing biomedical-MRC model. We extensively evaluate the performance of BioADAPT-MRC by comparing it with the best existing methods on three widely used benchmark biomedical-MRC datasets-BioASQ-7b, BioASQ-8b and BioASQ-9b. Our results suggest that without using any synthetic or human-annotated data from the biomedical domain, BioADAPT-MRC can achieve state-of-the-art performance on these datasets. AVAILABILITY AND IMPLEMENTATION: BioADAPT-MRC is freely available as an open-source project at https://github.com/mmahbub/BioADAPT-MRC. SUPPLEMENTARY INFORMATION: Supplementary data are available at Bioinformatics online.

Data Lakehouse is a new paradigm in data architectures that embodies and integrates already established concepts for the systematic management of disparate, large-scale data – a data lake for heterogeneous data management, use of open standards for high-performance querying, and systematic maintenance of the data "freshness". In addition to being a new concept, the data lakehouse is also still a conceptual construct. Many projects that use the lakehouse require maturing, empirical studies, and specific implementations. In this paper, we present our implementation of the data lakehouse concept in a biomedical research and health data analytics domain, and we discuss the implementation of some unique and novel features such as support for specialized access controls in support of HIPAA regulation and IRB protocols, and support for the FAIR standard.1

The study presented here focuses on performance characteristics and trade-offs associated with running machine-learning tasks in multi-GPU environments on both on-site cloud computing resources and commercial cloud services (Azure). Specifically, this study examines these tradeoffs by examining the performance of training and fine-tuning of transformer-based deep-learning (DL) networks on clinical notes and data, a task of critical importance in the medical domain. To this end, we perform DL-related experiments on the widely deployed NVIDIA V100 GPUs and on the newer A100 GPUs connected via NVLink or PCIe. This study analyzes the execution time of major operations to train DL models and investigate popular options to optimize each of them. We examine and present the findings on the impacts that various operations (e.g. data loading into GPUs, training, fine-tuning), optimizations, and system configurations (single vs. multi-GPU, NVLink vs. PCIe) have on the overall training performance.

Streaming data processing is an exercise in taming disorder: from oftentimes huge torrents of information, we hope to extract powerful and timely analyses. But when dealing with streaming data, the unbounded and temporally disordered nature of real-world streams introduces a critical challenge: how does one reason about the completeness of a stream that never ends? In this paper, we present a comprehensive definition and analysis of watermarks , a key tool for reasoning about temporal completeness in infinite streams. First, we describe what watermarks are and why they are important, highlighting how they address a suite of stream processing needs that are poorly served by eventually-consistent approaches: • Computing a single correct answer, as in notifications. • Reasoning about a lack of data, as in dip detection. • Performing non-incremental processing over temporal subsets of an infinite stream, as in statistical anomaly detection with cubic spline models. • Safely and punctually garbage collecting obsolete inputs and intermediate state. • Surfacing a reliable signal of overall pipeline health . Second, we describe, evaluate, and compare the semantically equivalent, but starkly different, watermark implementations in two modern stream processing engines: Apache Flink and Google Cloud Dataflow.

OBJECTIVE: Concept normalization, the task of linking phrases in text to concepts in an ontology, is useful for many downstream tasks including relation extraction, information retrieval, etc. We present a generate-and-rank concept normalization system based on our participation in the 2019 National NLP Clinical Challenges Shared Task Track 3 Concept Normalization. MATERIALS AND METHODS: The shared task provided 13 609 concept mentions drawn from 100 discharge summaries. We first design a sieve-based system that uses Lucene indices over the training data, Unified Medical Language System (UMLS) preferred terms, and UMLS synonyms to generate a list of possible concepts for each mention. We then design a listwise classifier based on the BERT (Bidirectional Encoder Representations from Transformers) neural network to rank the candidate concepts, integrating UMLS semantic types through a regularizer. RESULTS: Our generate-and-rank system was third of 33 in the competition, outperforming the candidate generator alone (81.66% vs 79.44%) and the previous state of the art (76.35%). During postevaluation, the model's accuracy was increased to 83.56% via improvements to how training data are generated from UMLS and incorporation of our UMLS semantic type regularizer. DISCUSSION: Analysis of the model shows that prioritizing UMLS preferred terms yields better performance, that the UMLS semantic type regularizer results in qualitatively better concept predictions, and that the model performs well even on concepts not seen during training. CONCLUSIONS: Our generate-and-rank framework for UMLS concept normalization integrates key UMLS features like preferred terms and semantic types with a neural network-based ranking model to accurately link phrases in text to UMLS concepts.

Real-time data analysis and management are increasingly critical for today's businesses. SQL is the de facto lingua franca for these endeavors, yet support for robust streaming analysis and management with SQL remains limited. Many approaches restrict semantics to a reduced subset of features and/or require a suite of non-standard constructs. Additionally, use of event timestamps to provide native support for analyzing events according to when they actually occurred is not pervasive, and often comes with important limitations. We present a three-part proposal for integrating robust streaming into SQL, namely: (1) time-varying relations as a foundation for classical tables as well as streaming data, (2) event time semantics, (3) a limited set of optional keyword extensions to control the materialization of time-varying query results. We show how with these minimal additions it is possible to utilize the complete suite of standard SQL semantics to perform robust stream processing. We motivate and illustrate these concepts using examples and describe lessons learned from implementations in Apache Calcite, Apache Flink, and Apache Beam. We conclude with syntax and semantics of a concrete proposal for extensions of the SQL standard and note further areas of exploration.

One of the key, emerging challenges that connects the "Big Data" and the AI domain is the availability of sufficient volumes of training data for AI/Machine Learning tasks. SynthNotes is a framework for generating standards-compliant, realistic mental health progress report notes at the very large, population-level scale, and in a strict privacy-preserving manner. Our framework, inspired by the needs to explore, evaluate, and train computational methods for the emerging mental health crisis in the US, is useful for benchmarking, optimization, and training of biomedical natural language processing, information extraction, and machine learning systems intended to operate at "Big Data" scale (billions of notes). The free text notes generated by SynthNotes are based on the literature and public statistical models allowing for realistic, natural language representation of a patient, and his or her mental health characteristics. Additionally, SynthNotes can partially simulate stylistic, grammatical, and expressive characteristics of a licensed mental health professional. SynthNotes is modular and flexible, allowing for representation of variety of conditions, incorporation of alternative foundational models, and parametrization of the variability of the structure, content, and size of the synthetically generated corpus. In this paper, we report on the initial use and performance characteristics of our SynthNotes framework and on the ongoing work for inclusion of content planning and deep learning-based generative methods trained on real data.

Apache Calcite is a foundational software framework that provides query processing, optimization, and query language support to many popular open-source data processing systems such as Apache Hive, Apache Storm, Apache Flink, Druid, and MapD. The goal of this paper is to formally introduce Calcite to the broader research community, brie y present its history, and describe its architecture, features, functionality, and patterns for adoption. Calcite's architecture consists of a modular and extensible query optimizer with hundreds of built-in optimization rules, a query processor capable of processing a variety of query languages, an adapter architecture designed for extensibility, and support for heterogeneous data models and stores (relational, semi-structured, streaming, and geospatial). This exible, embeddable, and extensible architecture is what makes Calcite an attractive choice for adoption in big-data frameworks. It is an active project that continues to introduce support for the new types of data sources, query languages, and approaches to query processing and optimization.

A Polystore concept, either in its orthodox form, or in a more general interpretation is relevant for a variety of the important, data-centric scenarios, most markedly precision medicine. As we continue to examine the feasibility and maturity of this concept, we are exploring both core and related technologies that are bringing us closer to a comprehensive solution of managing and efficient retrieval of the heterogeneous data. In this paper, we are presenting our exploration of Apache Calcite as a foundation for the more general polystore concept. To this end, we are discussing architectural characteristics of the platform, its performance, implementation, and usability trade-offs, and the readiness of the technology to serve as a polystore foundation. We examine Calcite against relational (SQL Server), array (TileDB), and text data engines (Elasticsearch).

The Polystore architecture revisits the federated approach to access and querying the standalone, independent databases in the uniform and optimized fashion, but this time in the context of heterogeneous data and specialized analyses. In light of this architectural philosophy, and in the light of the major data architecture development efforts at the US Department of Veterans Administration (VA), we discuss the need for the heterogeneous data store consisting of large relational data warehouse, an image and text datastore, and a peta-scale genomic repository. The VA's heterogeneous datastore would, to a larger or smaller degree, follow the architectural blueprint proposed by the polystore architecture. To this end, we discuss the current state of the data architecture at VA, architectural alternatives for development of the heterogeneous datastore, some relevant use cases, the anticipated challenges, and the drawbacks and benefits of adopting the polystore architecture.