Lightweight data management with dtool

Understanding machine learning applications in dementia research and clinical practice: a review for biomedical scientists and clinicians

Several (inter)national longitudinal dementia observational datasets encompassing demographic information, neuroimaging, biomarkers, neuropsychological evaluations, and muti-omics data, have ushered in a new era of potential for integrating machine learning (ML) into dementia research and clinical practice. ML, with its proficiency in handling multi-modal and high-dimensional data, has emerged as an innovative technique to facilitate early diagnosis, differential diagnosis, and to predict onset and progression of mild cognitive impairment and dementia. In this review, we evaluate current and potential applications of ML, including its history in dementia research, how it compares to traditional statistics, the types of datasets it uses and the general workflow. Moreover, we identify the technical barriers and challenges of ML implementations in clinical practice. Overall, this review provides a comprehensive understanding of ML with non-technical explanations for broader accessibility to biomedical scientists and clinicians.

dtool and dserver: A flexible ecosystem for findable data

Making data FAIR-findable, accessible, interoperable, reproducible-has become the recurring theme behind many research data management efforts. dtool is a lightweight data management tool that packages metadata with immutable data to promote accessibility, interoperability, and reproducibility. Each dataset is self-contained and does not require metadata to be stored in a centralised system. This decentralised approach means that finding datasets can be difficult. dtool's lookup server, short dserver, as defined by a REST API, makes dtool datasets findable, hence rendering the dtool ecosystem fit for a FAIR data management world. Its simplicity, modularity, accessibility and standardisation via API distinguish dtool and dserver from other solutions and enable it to serve as a common denominator for cross-disciplinary research data management. The dtool ecosystem bridges the gap between standardisation-free data management by individuals and FAIR platform solutions with rigid metadata requirements.

Molecular simulations of sliding on SDS surfactant films

We use molecular dynamics simulations to study the frictional response of monolayers of the anionic surfactant sodium dodecyl sulfate and hemicylindrical aggregates physisorbed on gold. Our simulations of a sliding spherical asperity reveal the following two friction regimes: at low loads, the films show Amonton's friction with a friction force that rises linearly with normal load, and at high loads, the friction force is independent of the load as long as no direct solid-solid contact occurs. The transition between these two regimes happens when a single molecular layer is confined in the gap between the sliding bodies. The friction force at high loads on a monolayer rises monotonically with film density and drops slightly with the transition to hemicylindrical aggregates. This monotonous increase of friction force is compatible with a traditional plowing model of sliding friction. At low loads, the friction coefficient reaches a minimum at the intermediate surface concentrations. We attribute this behavior to a competition between adhesive forces, repulsion of the compressed film, and the onset of plowing.

dtoolAI: Reproducibility for Deep Learning

Deep learning, a set of approaches using artificial neural networks, has generated rapid recent advancements in machine learning. Deep learning does, however, have the potential to reduce the reproducibility of scientific results. Model outputs are critically dependent on the data and processing approach used to initially generate the model, but this provenance information is usually lost during model training. To avoid a future reproducibility crisis, we need to improve our deep-learning model management. The FAIR principles for data stewardship and software/workflow implementation give excellent high-level guidance on ensuring effective reuse of data and software. We suggest some specific guidelines for the generation and use of deep-learning models in science and explain how these relate to the FAIR principles. We then present dtoolAI, a Python package that we have developed to implement these guidelines. The package implements automatic capture of provenance information during model training and simplifies model distribution.

Experimental Pipeline (Expipe): A Lightweight Data Management Platform to Simplify the Steps From Experiment to Data Analysis

As experimental neuroscience is moving toward more integrative approaches, with a variety of acquisition techniques covering multiple spatiotemporal scales, data management is becoming increasingly challenging for neuroscience laboratories. Often, datasets are too large to practically be stored on a laptop or a workstation. The ability to query metadata collections without retrieving complete datasets is therefore critical to efficiently perform new analyses and explore the data. At the same time, new experimental paradigms lead to constantly changing specifications for the metadata to be stored. Despite this, there is currently a serious lack of agile software tools for data management in neuroscience laboratories. To meet this need, we have developed Expipe, a lightweight data management framework that simplifies the steps from experiment to data analysis. Expipe provides the functionality to store and organize experimental data and metadata for easy retrieval in exploration and analysis throughout the experimental pipeline. It is flexible in terms of defining the metadata to store and aims to solve the storage and retrieval challenges of data/metadata due to ever changing experimental pipelines. Due to its simplicity and lightweight design, we envision Expipe as an easy-to-use data management solution for experimental laboratories, that can improve provenance, reproducibility, and sharing of scientific projects.