Details

Autor(en) / Beteiligte

• Gezsi, Andras
• Bruncsics, Bence
• Guta, Gabor
• Antal, Peter

Titel

Constructing a Quantitative Fusion Layer over the Semantic Level for Scalable Inference

Ist Teil von

• Bioinformatics and Biomedical Engineering, p.41-53

Ort / Verlag

Cham: Springer International Publishing

Link zum Volltext

Quelle

Alma/SFX Local Collection

Beschreibungen/Notizen

• We present a methodology and a corresponding system to bridge the gap between prioritization tools with fixed target and unrestricted semantic queries. We describe the advantages of an intermediate level of networks of similarities and relevances: (1) it is derived from raw, linked data (2) it ensures efficient inference over partial, inconsistent and noisy cross-domain, cross-species linked open data, (3) preserved transparency and decomposability of the inference allows semantic filters and preferences to control and focus of the inference, (4) high-dimensional, weakly significant evidences, such as overall summary statistics could also be used in the inference, (5) quantitative and rank based inference primitives can be defined, and (6) queries are unrestricted, e.g. prioritized variables, and (7) it allows wider access for non-technical experts. We provide a step-by-step guide for the methodology using a macular degeneration model, including drug, target and disease domains. The system and the model presented in the paper are available at bioinformatics.mit.bme.hu/QSF.