Abstract
Development of new molecular probes and drugs is a time intensive task that could be optimized using AI technologies. These AI technologies require well-annotated, searchable, findable, and reusable data that are following the FAIR principles. These data are required as the input for training these AI models as well as more standard high throughput screening campaigns as a method of prioritizing compounds. Currently, few resources are properly annotated using FAIR principles. without these resources annotated building AI models becomes sufficiently more difficult. The increased desire to develop these AI systems as well as inability to easily reuse old data has led to calls to make all data FAIR going forward.
In this dissertation I report my efforts to expand the tools required to annotate data to make it FAIR compliant through maintaining and expanding ontologies around assays and drug targets. Drug Target Ontology was developed and put into production with over six releases while I was the maintainer. BioAssay Ontology was expanded and converted to a more easily maintained format; this allowed release using the OntoJog tool developed by me. OntoloBridge is a yet to be released application to hopefully alleviate communication issues between researchers that generate data and need to annotate their data with the maintainers of ontologies. While in the lab I assisted in the final phases of release of the LINCS Data Portal and then designed the API and helped develop the database infrastructure for the LINCS Data Portal 2.0.