Searching for the urine osmolality surrogate: an automated machine learning approach

Clinical Chemistry and Laboratory Medicine (CCLM), Journal Year: 2022, Volume and Issue: 60(12), P. 1911 - 1920

Published: July 2, 2022

Automated machine learning (AutoML) tools can help clinical laboratory professionals to develop models. The objective of this study was a novel formula for the estimation urine osmolality using an AutoML tool and determine efficiency in setting.Three hundred routine urinalysis samples were used reference chemistry analysis. H2O engine completed development steps with minimum human intervention. Four feature groups created, which include different measurements according Boruta selection algorithm. Method comparison statistics including Spearman's correlation, Passing-Bablok regression analysis performed, Bland Altman plots created compare model predictions method. allowable bias (24.17%) from biological variation data as limit agreement.The developed total 183 ML Conductivity specific gravity had highest variable importance. Models that conductivity, gravity, other parameters R2 (0.70-0.83), 70-84% results within agreement.Combining urinary conductivity validated models yield promising surrogate. Additionally, facilitate cycle should be considered developing laboratories.

Language: Английский

---

Ten simple rules to make computable knowledge shareable and reusable

PLoS Computational Biology, Journal Year: 2024, Volume and Issue: 20(6), P. e1012179 - e1012179

Published: June 20, 2024

Computable biomedical knowledge (CBK) is: “the result of an analytic and/or deliberative process about human health, or affecting that is explicit, and therefore can be represented reasned upon using logic, formal standards, mathematical approaches.” Representing in a machine-interpretable, computable form increases its ability to discovered, accessed, understood, deployed. artifacts greatly advance the potential for implementation, reproducibility, extension by users, who may include practitioners, researchers, learners. Enriching help facilitate reuse translation into practice. Following examples 10 Simple Rules papers scientific code, software, applications, we present intended make shared more useful reusable. These rules are mainly researchers their teams have decided sharing important, wish go beyond simply describing results, algorithms, models via traditional publication pathways, want both research findings accessible, others use knowledge. roughly organized 3 categories: planning, engineering, documentation. Finally, while many following domains, these generalizable any domain.

Language: Английский

---

The Research Repository for Data and Diagnostics (R2D2): An Online Database Software System for High Performance Computing and Cloud-Based Satellite Data Assimilation Workflows

Lecture notes in computer science, Journal Year: 2024, Volume and Issue: unknown, P. 303 - 310

Published: Jan. 1, 2024

Language: Английский

---

Enabling Data-Driven Solubility Modeling at GSK: Enhancing Purge Predictions for Mutagenic Impurities

Organic Process Research & Development, Journal Year: 2024, Volume and Issue: unknown

Published: Nov. 11, 2024

In the pharmaceutical industry, solubility is a critical parameter influencing various stages of drug development, from early discovery to commercial manufacturing. This work showcases high-throughput screening workflow and describes steps required standardize curate data suitably allow automated flow. Using high-quality data, we developed quantitative structure–property relationship model using gradient boosting molecular descriptors, requiring only 2D structure generate predictions. The accuracy competitive with alternative approaches where additional physical not required. A key use case for predictions made in this way developing control strategies mutagenic impurities, allowing data-driven consistent method calculating contribution purge calculations. Further perspective given on future application as prediction algorithm approach methodologies supporting development general, highlighting potential federated learning which technological overcome barrier cross-industry sharing.

Language: Английский

---

Benefits and Limitations of Jupyter-based Scientific Web Applications

Published: Oct. 1, 2022

Scientists are increasingly interested in creating standalone web-applications as computational and data analysis tools. The authors have worked with several such research groups to design, develop, deploy web applications that based on Jupyter notebooks. One of the primary reasons among many use notebooks is fact inheriting these capable maintaining extending them. In this paper, we walk through design process for one application discuss development environments best suited notebook development. We then explore other where employ similar patterns. doing so, expound upon benefits, limitations, challenges Notebook-based provide a guide facilitators situations.

Language: Английский

---

Developing Data-Centric Clinical Laboratory Workflow Through the Use of Open-Source Tools

The Journal of Applied Laboratory Medicine, Journal Year: 2023, Volume and Issue: 8(1), P. 7 - 10

Published: Jan. 4, 2023

Journal Article Developing Data-Centric Clinical Laboratory Workflow Through the Use of Open-Source Tools Get access Deniz İlhan Topcu Medical Biochemistry Department, Başkent University Faculty Medicine, Ankara, Turkey Address correspondence to this author at: Department Biochemistry, Bağlıca Kampusü, Eskişehir yolu 20.km, Etimesgut, 06530, Turkey. E-mail [email protected]. https://orcid.org/0000-0002-1219-6368 Search for other works by on: Oxford Academic Google Scholar The Applied Volume 8, Issue 1, January 2023, Pages 7–10, https://doi.org/10.1093/jalm/jfac110 Published: 04 2023 history Received: 09 October 2022 Accepted: 24

Language: Английский

---

Enhancing the accessibility and interactions of regional hydrologic projections for water managers

Environmental Modelling & Software, Journal Year: 2023, Volume and Issue: 167, P. 105763 - 105763

Published: June 24, 2023

Language: Английский

---