INTEGRATING MORPHOSOURCE, A COMMUNITY DIGITAL REPOSITORY FOR 3D SPECIMEN DATA, WITH MORPHOMETRIC ANALYSIS PIPELINES

Over the past several decades, the use of 3D data has become increasingly common in paleontological research, and techniques leveraging 3D data offer great promise for scientific advancement. This purpose can be better achieved by increasing the availability of data through sharing and archiving, as well as improving the accessibility of computational morphometric workflows for deriving insights from available data. We discuss efforts to accomplish both of these goals. MorphoSource is a publicly accessible digital repository of 3D data that has existed since 2013. At the time of writing, it holds over 27,000 3D datasets representing almost 2,000 taxa. Users can upload and download high fidelity 3D renderings of specimens derived from a variety of scanning modalities. Specimens (ideally vouchered in a physical collection) comprise the basic unit of organization. Users who upload data maintain control over access to that data, and DOI identifiers can be generated that allow uploaders to be cited for usage of data when they share it. Shared data can be easily searched and downloaded. Moving forward, we propose to integrate the MorphoSource repository with a variety of analytical infrastructures, allowing the data on MorphoSource to serve different communities and questions. Specifically we plan to (a) fortify widely relied upon 3D geometric morphometric approaches based on analysis of user-placed landmarks, by developing extensions to the open source volume software 3D Slicer that allow users to query and download data from MorphoSource within the application; and (b) to explore new paradigms of shape analysis, by linking MorphoSource to the Shape Analysis for Phenomics from Imaging Data (SAPID) software, hosted as a “Science Gateway” using NSF XSEDE computational resources. The latter integration will allow cloud-based morphometric analyses to be carried out on the MorphoSource website or locally using data hosted on MorphoSource servers. The joining of massive numbers of 3D digital specimens with state-of-the-art analytical pipelines and powerful computing resources is essential for transforming comparative morphology into a ‘big data’ science with a pace and power on par with genomics.