From iDigBio
Jump to: navigation, search

MorphoSource: The first open access digital archive for high fidelity 3D data on morphological phenomes

Project Summary


This project--led by a multi-institutional team of comparative biologists, library digital archivists, and computer scientists--envisions the development of novel, archive-quality cyber infrastructure for the secure storage of large volumes of 3D data on organismal anatomical structure (morphological phenomes) in a sustainable and scalable manner. The proof-of-concept (called MorphoSource) is an open-access archive and the only one of its kind; it is a GenBank-style repository that supports 3D digital renderings of objects catalogued in scientific collections of museums and universities.

The major project goal is to transition MorphoSource to a full-scale archive that can support research and education on a national level. This primarily involves refactoring MorphoSource from a custom web application to a widely used digital repository platform. This task will enhance scalability, sustainability, security, and data preservation potential by eventually allowing a) transferrable institutional governance, b) mass ingest of large numbers of datasets, c) easy addition of partner institutions to increase capacity, d) increases in the copy number and geographic distribution of repository content, e) state-of-the-art blockades against bit-rot and software-rot, and f) development of infrastructure for cost recovery. In addition, this project will enhance long-term impact and preservation potential of 3D data of many previously funded NSF projects by migrating legacy data from Pennsylvania State University and Duke University to MorphoSource. Finally, the team will apply MorphoSource in support of public education and develop appropriate features.

Whereas all published genomic datasets can be easily accessed and re-analyzed through online repositories, without MorphoSource, no such venue is available for 3D data on morphological phenotypes. Thus, MorphoSource is an essential component for synchronizing analysis of genomic and phenomic data—one of the biggest challenges for biological sciences. 3D digital phenotypes on MorphoSource can be analyzed by automated, high throughput quantifications of complex shape and/or machine learning approaches to discover patterns of covariation between morphology and other aspects of the phenome, genome, and environment.

Transitioning MorphoSource to a more robust and scalable repository infrastructure is essential for fostering a full transformation in the culture of data sharing and for enabling a much larger community of researchers to benefit from morphological data. A primary obstacle for building an open access morphology database has been designing a storage framework that could grow easily and robustly with demand. The PIs detail a plan for developing such infrastructure that employs Fedora, the leader in open source repository development. This initiative will extend the latest version of Fedora in novel and impactful ways to build storage and eventually allow object node management across multiple institutions.

When 3D data are served on MorphoSource and made universally accessible and reusable, the actual and demonstrable research and education value of the data and the physical specimens the data represent are magnified in an unprecedented way. Stakeholders (data authors, museums and/or funding agencies) can export usage statistics or check citations of data DOIs, which constitute powerful evidence of scientific value and broader impacts. The need for physical access or means to digitize important collections will be reduced, allowing researchers and funders to focus their time and money on experimental design and data analysis. Finally, a well-populated MorphoSource brings data on unity of life and biological diversity to the public in an easily accessible and broadly available format. Already teachers and students from around the world have downloaded and use thousands of MorphoSource datasets. Additionally, the PIs and collaborators will develop and demonstrate outreach tools that build on the initial success of MorphoSource bringing the raw data on natural history to k-12 classrooms in a way never before possible.

Project Websites & Social Media

MorphoSource (on-line depository for digital anatomical data):

MorphoSource on twitter:

MorphoSource YouTube Channel:

Normalized terms for 3D data modalities:

Project Leadership

Project sponsor: Duke Univ. (NSF Award DBI-1661386)

Principal Investigator (PI): Doug Boyer

Project collaborators

Duke University

Doug M. Boyer, lead PI, Duke Univ., Dept. of Evolutionary Anthropology (NSF Award DBI-1661386)
Timothy M. McGeary, CoPI, Duke Univ. Libraries
Sayan Mukherjee,CoPI, Duke Univ., Dept. of Statistics and Computer Science
Senior personnel
Dr. Julia M. Winchester – Lead Developer and Product Manager, Dept. of Evolutionary Anthropology
Mackenzie Shepard – Lab Manager and Digital Technician, Dept. of Evolutionary Anthropology
Simon Choy – Developer, Duke Univ. Libraries
Jocelyn Triplett – Developer, Duke Univ. Libraries

Pennsylvania State University (Penn State)

Timothy Ryan, lead PI, Penn State Univ., Dept of Anthropology (NSF Award DBI-1661132)
Kathleen Hill, CoPI, Penn State College of Education, Center for Science And The Schools (CSATS)
Senior personnel
Joshua Wisor – Digital data technician, Dept of Anthropology
Taylor Wood – Research Assistant, Penn State College of Education, CSATS
Amber Cesare – Research Assistant, Penn State College of Education, CSATS