Interoperability

Public by Private

Data protected by privacy or IP issues need not remain isolated from public data in G.E.M.S. Since the data owners completely control who they share their data and tools with, they can freely link data across the public / private divide with trusted partners on the platform.

G x E x M x S

Ordinarily data is siloed in individual Genomics, Environment, Management (Agronomy) and Socioeconomic communities, with little consideration of leveraging data across disciplines to give a more complete picture of the agricultural system under study. G.E.M.S facilitates integration of data across these data domains, primarily through spatial and temporal co-localization, metadata enhancement, and regularization of vocabulary.

Space by Time

G.E.M.S is careful not to overlook relationships implicit in seemingly unrelated data. Two of the most powerful inferential links include spatial and temporal co-localization. High-throughput phenotypic data gathered from ongoing field trials can be richly enhanced with high-resolution weather data, satellite imagery and soil sensor data when all of those streams of data were gathered at the same time and place.

Ontologies

A big barrier to incorporating legacy data or to analyzing your research data in the context of that gathered by another group are differences in vocabulary, units of measurement, and experimental protocols. We have been working very closely with world leaders in agricultural ontologies to create a data input system that allows you to map spreadsheets of data to the most appropriate ontological terms. This provides a common framework for each research project to talk to the others.

Cleaning and Imputation

Garbage in, garbage out is not for us! There are lots of numbers masquerading as data. We have invested considerable amount of resources into a suite of tools that enables researchers to remove spelling errors, consolidate synonymous terms, impute missing data, and flag outliers in both spreadsheets and spatial coordinates. Hence data can be cleaned upon ingestion, before it is registered in GEMShare keeping our shared data repositories clean.