Phenotype Annotation at TAIR

Phenotype Annotation at TAIR

Tanya Berardini

Current Status

• Phenotypes are associated to organisms (germplasms) and linked to references

• Curation is mainly done in unstructured full text

• Pilot study of structured annotation completed

• PaTO not yet implemented

papillae:abnormalstyle:abnormalseptum development:incompleteleaf:serrate

Same challenges as other databases

• Large corpus of literature

• Set number of curators with many diverse responsibilities

• Different ways of paper prioritization

Capturing phenotypes from the literature

• Single paper, describing single aspect of phenotype

• Single paper, describing multiple aspects of the organism’s phenotype

• Multiple papers describing different aspects of the organism’s phenotype

• Multiple papers, describing the same phenotype

Storing complex annotations in the database

• Right now, GO as a model for assocations:– 1 association = 1 gene product + 1 CV term

• How do we deal with multiple ontologies in one association?– 1 association = 1 germplasm + 1 anatomy term

+ 1 PATO term (+ others)