Dataset Information

Desiderata for sharable computable biomedical knowledge for learning health systems.

ABSTRACT: In this commentary, we work out the specific desired functions required for sharing knowledge objects (based on statistical models) presumably to be used for clinical decision support derived from a learning health system, and, in so doing, discuss the implications for novel knowledge architectures. We will demonstrate how decision models, implemented as influence diagrams, satisfy the desiderata. The desiderata include locally validate discrimination, locally validate calibration, locally recalculate thresholds by incorporating local preferences, provide explanation, enable monitoring, enable debiasing, account for generalizability, account for semantic uncertainty, shall be findable, and others as necessary and proper. We demonstrate how formal decision models, especially when implemented as influence diagrams based on Bayesian networks, support both the knowledge artifact itself (the "primary decision") and the "meta-decision" of whether to deploy the knowledge artifact. We close with a research and development agenda to put this framework into place.

SUBMITTER: Lehmann HP

PROVIDER: S-EPMC6508769 | biostudies-literature | 2018 Oct

REPOSITORIES: biostudies-literature

ACCESS DATA

Publications

Desiderata for sharable computable biomedical knowledge for learning health systems.

Lehmann Harold P HP Downs Stephen M SM

Learning health systems 20180803 4

In this commentary, we work out the specific desired functions required for sharing knowledge objects (based on statistical models) presumably to be used for clinical decision support derived from a learning health system, and, in so doing, discuss the implications for novel knowledge architectures. We will demonstrate how decision models, implemented as influence diagrams, satisfy the desiderata. The desiderata include locally validate discrimination, locally validate calibration, locally recal ...[more]

PMID: 31245589

Similar Datasets

Project description:IntroductionWhile data repositories are well-established in clinical and research enterprises, knowledge repositories with shareable computable biomedical knowledge (CBK) are relatively new entities to the digital health ecosystem. Trustworthy knowledge repositories are necessary for learning health systems, but the policies, standards, and practices to promote trustworthy CBK artifacts and methods to share, and safely and effectively use them are not well studied.MethodsWe conducted an online survey of 24 organizations in the United States known to be involved in the development or deployment of CBK. The aim of the survey was to assess the current policies and practices governing these repositories and to identify best practices. Descriptive statistics methods were applied to data from 13 responding organizations, to identify common practices and policies instantiating the TRUST principles of Transparency, Responsibility, User Focus, Sustainability, and Technology.ResultsAll 13 respondents indicated to different degrees adherence to policies that convey TRUST. Transparency is conveyed by having policies pertaining to provenance, credentialed contributors, and provision of metadata. Repositories provide knowledge in machine-readable formats, include implementation guidelines, and adhere to standards to convey Responsibility. Repositories report having Technology functions that enable end-users to verify, search, and filter for knowledge products. Less common TRUST practices are User Focused procedures that enable consumers to know about user licensing requirements or query the use of knowledge artifacts. Related to Sustainability, less than a majority post describe their sustainability plans. Few organizations publicly describe whether patients play any role in their decision-making.ConclusionIt is essential that knowledge repositories identify and apply a baseline set of criteria to lay a robust foundation for their trustworthiness leading to optimum uptake, and safe, reliable, and effective use to promote sharing of CBK. Identifying current practices suggests a set of desiderata for the CBK ecosystem in its continued evolution.

Project description:Knowledge graphs (KGs) are an important tool for representing complex relationships between entities in the biomedical domain. Several methods have been proposed for learning embeddings that can be used to predict new links in such graphs. Some methods ignore valuable attribute data associated with entities in biomedical KGs, such as protein sequences, or molecular graphs. Other works incorporate such data, but assume that entities can be represented with the same data modality. This is not always the case for biomedical KGs, where entities exhibit heterogeneous modalities that are central to their representation in the subject domain. We aim to understand how to incorporate multimodal data into biomedical KG embeddings, and analyze the resulting performance in comparison with traditional methods. We propose a modular framework for learning embeddings in KGs with entity attributes, that allows encoding attribute data of different modalities while also supporting entities with missing attributes. We additionally propose an efficient pretraining strategy for reducing the required training runtime. We train models using a biomedical KG containing approximately 2 million triples, and evaluate the performance of the resulting entity embeddings on the tasks of link prediction, and drug-protein interaction prediction, comparing against methods that do not take attribute data into account. In the standard link prediction evaluation, the proposed method results in competitive, yet lower performance than baselines that do not use attribute data. When evaluated in the task of drug-protein interaction prediction, the method compares favorably with the baselines. Further analyses show that incorporating attribute data does outperform baselines over entities below a certain node degree, comprising approximately 75% of the diseases in the graph. We also observe that optimizing attribute encoders is a challenging task that increases optimization costs. Our proposed pretraining strategy yields significantly higher performance while reducing the required training runtime. BioBLP allows to investigate different ways of incorporating multimodal biomedical data for learning representations in KGs. With a particular implementation, we find that incorporating attribute data does not consistently outperform baselines, but improvements are obtained on a comparatively large subset of entities below a specific node-degree. Our results indicate a potential for improved performance in scientific discovery tasks where understudied areas of the KG would benefit from link prediction methods.

Dataset Information

Desiderata for sharable computable biomedical knowledge for learning health systems.

Publications

Desiderata for sharable computable biomedical knowledge for learning health systems.

Similar Datasets

OmicsDI is part of the ELIXIR infrastructure

Tweets