Leveraging ontologies to transform insect taxonomy and phylogenetics

1. Leveraging ontologies to transform insect taxonomy and phylogenetics Andy Deans North Carolina State University John Hallmén

2. Acknowledgments • funding: • NSF Advances in Biological Informatics • Morphbank (NSF DBI-0446224) • HymAToL (NSF EF-0337220) • PEET: Monographic research on parasitic Hymenoptera (NSF DEB-0328922) • intellect and enthusiasm: • István Mikó, Katja Seltmann, Matt Yoder (NSCU) • Fredrik Ronquist (NRM) • Jim Balhoff, Hilmar Lapp, Todd Vision, Wasila Dahdul (NESCent) • Paula Mabee (USD) • Anne Maglia (MUS & T) • All the contributors! (especially the International Society of Hymenopterists) Hymenoptera images: http://www.flickr.com/photos/orionmystery/1777817613 http://www.flickr.com/photos/leapfrog_photo/2893205919/ http://www.flickr.com/photos/sanmartin/2320291727/ http://www.flickr.com/photos/chi-liu/400478069/ http://www.flickr.com/photos/mcduck/2307414339/ http://www.flickr.com/photos/johnhallmen/3021409417/

3. Part I Anatomy in Hymenoptera systematics Part II Hymenoptera Anatomy Ontology Part III How to build this resource Part IV How to use this resource The End Questions?

4. Hymenoptera >115,000 spp. sawflies ants bees social wasps parasitic wasps sanmartin orionmystery Gary McDonald chi-liu Art

5. >115,000 species descriptions Head transverse to subquadrate in dorsal view; hyperoccipital carina absent; occipital carina complete, crenulate, closely approximated to foramen magnum dorsally; lateral ocellus contiguous with inner orbit; eye glabrous or sparsely setose; frons nearly flat, antennal scrobe not developed; inner orbits distinctly diverging ventrally.

6. 7,150 functional morphology papers The entry into the wood can be done by first retracting the abdominalsegments into their normal position and then by pressure of the whole abdomen; in most cases observed, the stylus comes into contact with the wood. (Le Lannic & Nénon 1999) 7,900 morphological phylogenetics papers 1. Shape of tentorial bridge: (0) straight (1) distinctly arched. 2. Corpotendon (anterior process from upper tentorial bridge forming the tendon of the posterior contractor of the pharynx): (0) absent; (1) present.

7. 202 terms We have NO single anatomical reference ad hoc glossaries

8. Terminological Nightmare

9. Homonyms paramere

10. Synonyms unguis tarsal claw pretarsal claw

11. Other Terminological Problems alitrunk uninformative terms (=to be discouraged?) and taxon-specific terms gaster “lost” terms thigmomere thigmus thigmochore

12. The ever increasing confusion in the application of anatomical terminology in entomology, is rapidly producing an absolutely intolerable state of affairs... Such chaotic ... needless confusion would not ... be tolerated in any other branch of research Grimaldi & Engel 2005 - Guy C. Crampton (1915)

13. Serious Implications for Systematics 25 taxonomic revisions from 5 lineages: 5 “Symphyta” 5 Ichneumonoidea 5 Evaniomorpha + misc. 5 Aculeata 5 Chalcidoidea

14. Serious Implications for Systematics 296 total anatomical terms (mean=43): • 28 common across all • 155 lineage-specific • 3-5 overlapped within a lineage • 86 author-specific

15. Serious Implications for Systematics non-homologous characters duplicated characters miscommunication wasted effort

16. Serious Implications for Genomics mutant phenotype description gene expression annotation “stumpy” = abdomen shortened “hunchback” = thoracic segments compressed “glass” = eye facets poorly differentiated and number reduced “short wings” = small mesothoracic wings; metathoracic wings project out from body

17. Part I Anatomy in Hymenoptera systematics Part IIHymenoptera Anatomy Ontology Part III How to build this resource Part IV How to use this resource The End Questions?

18. What is an ontology? ontology ≠ ontogeny The history of structural change in a unity, which can be a cell, an organism, or a society of organisms, without the loss of the organization that allows that unity to exist (Maturana and Varela, 1987).

19. a) textualdefinition b) annotation anatomical structures Hymenoptera anatomy is_a part_of ontology = Formal representation of concepts within a domain and the relationships between those concepts. (in Computer Science)

20. harpe The sclerite on the external male genitalia that is connected basally to the gonostipes via cojunctiva and the proximal and distal gonostipes-harpe muscles. Formal representation of concepts within a domain and the relationships between those concepts. harpe is_asclerite harpe part_ofexternal male genitalia parameresynonymharpe palettesynonymharpegonosquamasynonymharpe

21. genus-differentia = 1. definiendum: the term being defined 2. genus: the broader category for that term (the difiniendum’s parent) 3. differentia: how that term differs from the genus’s other children B is an A that X leg is a segmented appendage that is involved in locomotion

22. mesoscutum fore wing costal vein fore wing pterostigma tegula head eye (or compound eye) tergite flagellomere clypeus sternite fore leg tibia hind leg tibia mesopleuron fore leg basitarsus hind leg femur tarsal claw Hymenoptera anatomy... John Hallmén

23. Other relevant ontologies Gene Ontology eye pigment biosynthetic process "The chemical reactions and pathways resulting in the formation of eye pigments, any general or particular coloring matter in living organisms, found or utilized in the eye." (GOC:ai) eye pigment anabolismsynonymeye pigment biosynthetic process eye pigment biosynthesissynonymeye pigment biosynthetic process eye pigment formationsynonymeye pigment biosynthetic process eye pigment synthesissynonymeye pigment biosynthetic process eye pigment biosynthetic processis_aeye pigment metabolic process eye pigment biosynthetic processis_apigment biosynthetic process

24. Other relevant ontologies Units of Measurement millimeter "A length unit which is equal to one thousandth of a meter or 10^[-3] m." (http://physics.nist.gov/cuu/Units/) millimeteris_alength unit mmsynonym millimeter

25. Other relevant ontologies Phenotypic Quality (PATO) yellow "A color hue with medium wavelength of that portion of the visible spectrum lying between orange and green, evoked in the human observer by radiant energy with wavelengths of approximately 570 to 590 nanometers." (http://dictionary.reference.com/) yellowis_acolor

26. Other relevant ontologies Spatial dorsal to dorsal toinverse_ofventral to superior tosynonymdorsal to

27. Other relevant ontologies Taxonomic Apis Apis is_aApini Apis andreniformisis_aApis Apis floreais_aApis Apis dorsatais_aApis Apis ceranais_aApis Apis koschevnikoviis_aApis Apis melliferais_aApis Apis nigrocinctais_aApis

28. Why do this? 1. standardize our vocabulary for future publications 2. extract information from publications/databases

29. partonomy!

30. Part I Anatomy in Hymenoptera systematics Part II Hymenoptera Anatomy Ontology Part III How to build this resource Part IV How to use this resource The End Questions?

31. getting started...

32. backend data dump .obo [Term] id: 587 name: scrobal groove def: "A horizontal groove on the mesopleuron that may be continuous with the episternal groove anteriorly and ends at the pleural grove posteriorly” ref: "Goulet H, Huber JT, 1993. Hymenoptera of the World: An Identification Guide to Families. Research Branch, Agriculture Canada Publication 1894/E., Ottawa, ON. 668 pp.” relationship: part_of 157 ! mesopleuron relationship: is_a 138 ! groove CS colleagues: - unsupervised machine learning - semantic image searching OBO Foundry mx

34. Accessioning data - manually

35. Accessioning data - text mining and extraction Fisher BL, Smith MA. 2008. A revision of Malagasy species of Anochetus Mayr and Odontomachus Latreille (Hymenoptera: Formicidae). PLoS ONE 3(5): e1787 doi:10.1371/journal.pone.0001787

37. Reaccessioning data - tagging

38. Where we are now... 2,214 terms 2,234 relationships 13 contributors (of 40 added to the project) • homonyms: • e.g., anellus, speculum, pedicel, gaster, face, stigma, disc, metapleural triangle, uncus, paramere...etc. • chaotic character systems: • e.g., propodeal ridges, pronotal ridges, glands, occipital carinae, cuticular patches, male and female genitalia, thoracic musculature

39. Part I Anatomy in Hymenoptera systematics Part II Hymenoptera Anatomy Ontology Part III How to build this resource Part IV How to use this resource The End Questions?

40. 1. Text Mark-up • quality control • provide definitions

41. homonym! proofing tool http://hymglossary.tamu.edu/

42. http://evanioidea.info

43. Web-accessible taxon descriptions • high-lighting for definitions • and feedback

44. 1. Text Mark-up • quality control • provide definitions • in concert with other ontologies (future)

45. distance from the carina posterior to the mesoscutellum to the process dorsal to the propodeal foramen = 0.11 mm distance from the carinaposterior to the mesoscutellum to the processdorsal to the propodeal foramen = 0.11 mm PATO spatial HAO units of measurement

46. 2. Search Algorithm (future) • exploit the logic for efficient queries • diagnostic tools

48. Search for - “claw” “auxilia” “empodium” “empodia” “planta” “plantae” “pretarsus” “pretarsi” “orbicula” “orbiculae” “arolium” “pulvillum” “pulvilla” “unguifer” “manubrium” “ungues” “unguis” “arcus” “dorsal plate”

49. OR “pretarsus” [] include related terms

50. Flagellomere 1: (0) without multiporous plate sensilla (1) with multiporous plate sensilla