PurposeThe Surrey (British Columbia, Canada) fire department has an annual cycle for hiring full-time firefighters. This paper optimizes the timing of the annual hiring period. A key issue is handling workplace absences, which can be covered by overtime cost or full-time hires.Design/methodology/approachShort-term and long-term absences patterns are analyzed according to season and age cohorts of the firefighters. These are then used in both an explanatory and time series model to predict future absences. The hiring schedule is optimized based on these predictions and additional constraints.FindingsThe current practice fares well in the analysis. For the time period studied, moving to earlier hiring dates appears beneficial. This analysis is robust with respect to various assumptions.Originality/valueThis is a case study where analytic techniques and machine learning are applied to an organizational practice that is not commonly analyzed. In this case, the previous method was not much worse than the optimized solution. The techniques used are quite general and can be applied to various organizational decision problems.
Background: The field of cancer diagnosis and therapy is being empowered by Next Generation Sequencing (NGS) technology in recent years. The correlations among cancer, biomarker and drug are usually analyzed and used to guide the personalized drug recommendation for patients. However, the accurate correlation analysis faces two major challenges. First, large amount of existing variants carry the high variability in pathogenicity and clinical relevance. Second, the knowledge concerning the correlations is not easily accessible in the case of rare variants. The quantity of currently known variants, and dramatic increase of newly identified variants have showed the need for building up a knowledge base, for interpreting correlations among cancer, biomarker and drug. Method: To make the knowledge base accurately maintainable and friendly usable, we designed knowledge structures (or ontologies), and implemented an ontology based Clinical Knowledge Base (CKB). CKB consists of evidences from three different aspects, including cancer (i.e. disease), biomarker (genes and their corresponding variants), and drug. The inclusion of Gene Ontology (GO), Sequence Ontology (SO), Disease Ontology (DO) and KEGG Pathway Ontology (PO) not only describes the individual entities and the relationships between them, but also facilitates the discovery of correlations for potential biomarkers through ontology searching mechanism. CKB was implemented in Java and MySQL, and it provided a web based UI to input, analyze, relate and retrieve the data. Results: CKB supports effective retrieval of comprehensive information of cancer, biomarker and drug, and it highlights the sophisticated relationship among the respective entities. By inputting a biomarker of interest, evidences associated with the relevant genes/variants and relevant cancer types will be displayed. Evidences will be sorted by the significance levels. The levels include FDA/NMPA approved drugs, NCCN guidelines, clinical trials, clinical case report, pre-clinical evidence, and others. Furthermore, for the advanced users, the correlations between biomarker and cancer type can be explored for the research perspective of umbrella or basket clinical trials. CKB is open to public for research use at ckb.sodayun.com. Conclusions: Ontology is a powerful modeling methodology for medical knowledge. To manage the exponentially increasing NGS variants and their correlation with cancer and drug, ontology based knowledge base like CKB will play a significant role to help provide accurate correlation interpretation. It will be valuable for researchers and clinicians to better understand the correlations, in order to design new diagnostic assay or prescribe therapeutic regimens for their patients. For the more usage of CKB, more research is required, for example, how to embed CKB ontologies into a NGS pipeline. Citation Format: Jing Ma, Yizhou Ye, Hua Dong, Bingru Sun, Bolong He, Hank Yang. CKB: a Clinical Knowledge Base for interpreting correlations among cancer, biomarker and drug [abstract]. In: Proceedings of the Annual Meeting of the American Association for Cancer Research 2020; 2020 Apr 27-28 and Jun 22-24. Philadelphia (PA): AACR; Cancer Res 2020;80(16 Suppl):Abstract nr 3218.
scite is a Brooklyn-based organization that helps researchers better discover and understand research articles through Smart Citations–citations that display the context of the citation and describe whether the article provides supporting or contrasting evidence. scite is used by students and researchers from around the world and is funded in part by the National Science Foundation and the National Institute on Drug Abuse of the National Institutes of Health.
hi@scite.ai
10624 S. Eastern Ave., Ste. A-614
Henderson, NV 89052, USA
Copyright © 2024 scite LLC. All rights reserved.
Made with 💙 for researchers
Part of the Research Solutions Family.