Management Plan and Timeline
From HAO Wiki
Personnel
The current Hymenoptera Anatomy Ontology working group consists primarily of morphologists and systematists. We propose creating a "board of advisors" that includes a broader array of experts (e.g., computer scientists, ontologists, Hymenoptera genomics people) who will 1) provide critical input on the infrastructure that the HO is based on and 2) help to integrate the HAO into the larger community. We have identified potential board members and expect this formulation to be finalized prior to January 2009.
The short list of official investigators on this project betrays the fact that the HAO is a fully integrated, comprehensively collaborative effort that remains well connected to the Hymenoptera systematics (Ronquist, Heraty, Austin, La Salle, Schauff letters) and morphology (Vilhelmsen) communities, labs researching gene expression in the honey bee (Schatz letter) and Nasonia (Werren and Desjardins letter), ontologists from other domains (Vision, Maglia, Wiegmann and Yeates letters; Mabee in litt. regarding RCN), other informatics efforts (EOL, Cui, Agosti, Miranker and Humphries letters). We have assembled a dynamic team of specialists with a history of working well together:
Andrew Deans (PI), Hymenoptera systematist, morphologist.— The HAO was initiated by Deans and Ronquist (2006), and the PI serves as the spokesperson and primary driver for the initial surge. Deans serves as director of the NCSU Insect Museum, an institutional member of TDWG and an entity actively engaged in phyloinformatics related to specimens and taxon names. The Deans lab actively engages in systematics research on the major (and distantly related) hymenopteran lineages Evanioidea and Ceraphronoidea; the primary data type for these projects is morphology. Members of the Deans lab will serve as major contributors to and testers of the HO and its associated tools. PI Deans is an author of mx.
Matt Yoder (researcher), Hymenoptera systematist, morphologist, programmer.— Yoder has a cross-domain skill set that links informatics-related endeavors with empirical approaches. He is the lead developer for mx, which is one of the core tools in the proposed project. Matt also provides practical expertise on a large but relatively poorly known group of Hymenoptera, the Diapriidae, he is the only recent PhD in systematics of this group. Yoder will be the lead developer of the underlying database and the front ends required for data curation. He will further help develop (in conjunction with the Programmer) the automated approaches to data gathering, refinement and presentation. Yoder will work in close conjunction with Mikó to develop efficient tools from the users perspective.
Katja Seltmann (project manager).— Computer specialist, biological data archivist. A project of this magnitude requires an experienced manager and promoter. Katja Seltmann will serve in this role and brings with her experience in PHP, SQL, XML, an extensive knowledge of and connection to a critical resource (Morphbank), and strong background knowledge of Hymenoptera. Katja maintains invaluable connections with other informatics projects that will help extend our efforts to reach the broadest array of potential users possible. Katja will act as the point person with respect to community related activities, including evangelism, workshop development, helpdesk, feedback and testing, and quality assurance.
István Mikó (research associate), Hymenoptera morphologist.— Miko's anatomical knowledge spans the order Hymenoptera, and he continuously refines his skills in specimen preparation and imaging, image manipulation and annotation. Mikó harbors a penchant for proper terminology and is developing an interest in evolutionary development research as it pertains to establishing homologies. Mikó will focus on refining relationships within the HAO, identifying potential "preferred" terms and typifying terms with images and annotations in Morphbank.
Programmer (research associate).— We will recruit a programmer with 3-5 (5 preferred, graduate work taken on par) years experience in Ruby, SQL, Apache, Perl, and RESTful web services. PHP, Javascript, XML, and Ajax-related skills are a bonus. This individual will be hired to program lower level functionality and linkages (primarily Web services and related) between existing and newly created software. Applicants with a background in evolutionary (organismal) biology will be particularly competitive, though this will not be required. The programmer will focus on developing web-services pertaining to the HO, but will also be involved in developing and implementing the frameworks for automated data gathering and presentation.
Matt Bertone (postdoc) Diptera morphologist, software tester.— Bertone recently earned his Ph.D. in dipteran phylogenetics (focused on brachyceran taxa) but is broadly trained in Insecta systematics. We project requires a specialist of his caliber and breadth on order to integrate and map the only other insect-focused ontologies, both of which are dipteran: the Mosquito and Drosophila Gross Anatomy ontologies. Terry Catapano (consultant).— Terry is a library scientist with expertise in XML mark-up and semantics. Through his interaction with Plazi he provides a testbed corpus of TaxonX/XML encoded treatments of ants (Formicidae). He will work to enhance Plazi’s mark-up to explicitly relate the encoded text to the controlled terms in the ontology so as to expose the data for efficient retrieval and integration in a networked environment. Plazi will aim to investigate practices and deliver in Plazi web services XML (in TaxonX, SDD, and the NLM/NCBI DTD) and RDF (in SPM) utilizing the HAO. Graduate students.— At least two graduate students will be employed to contribute terms, use the ontology-enhanced tools (mx and others) for species descriptions and morphological phylogenetics.
Undergraduate assistant.— One undergraduate student will be hired to OCR select literature and to assist in imaging and other duties.
Sustainability Plan
As knowledge accumulates and more taxa are described and more genomes get sequenced the Hymenptera Anatomy Ontology will necessarily remain a dynamic entity. By exploiting existing resources (OBO Foundry, SourceForge, Google Code, GitHub, our project and lab websites) we will ensure permanent availability of the ontology itself and of the tools we develop as part of this project. Any decisions regarding the direction and long term preservation of the HO DB will be made by the governing board, however, as all our products, including our data, will be shared under open source Creative Common's licenses, there will be no restrictions on how copies of the data or applications can be used in the future. Because the database management of the HO will be tied to the mx software, which has been adopted by use for several members of the Hymenoptera community, there is an additional built in failsafe for longterm maintenance of the database version of the HAO. In addition, thanks to close societal ties with the ISH, we have a institution with a vested interest in keeping the HAO "alive".
Workflow
Our anticipated workflow is outlined in Fig. 1. Ontology development is an iterative process, and we have partnered with several other projects (Hong Cui (PI) and Bob Morris's "Fine-Grained Semantic Markup of Descriptive Data for Knowledge Applications in Biodiversity Domains" ABI, Donat Agosti at Plazi.org, Anne Thessen and David Patterson at the Encyclopedia of Life) that can provide feedback regarding the logic and completeness of the HAO.
Fig. 1 A. Overview of workflow interactions. Lines of larger green circles indicate personal interactions, those of smaller orange circles data. Rounded green boxes are applications, blue ovals are Ontologies, yellow circles are groups of personnel. B. Simplified table structure presently used to store the HO. Tags and figures can be applied to any object with an id. We plan to develop a "TagEngine" which will display Tagged objects based on context and keyword. Tags can be used to create subsets of the HO, which can then, for instance, be exported to other ontologies. Morphbank images are linked to any object with an id through a Figure record. Orange lines = data flow, green lines = human interaction.
