This is just a repo to save code-work implementing a p-value aggregation of sleuth differential expression p-values of individual transcripts over GO terms. So this is work in progress, but should work on sleuth results in RDS format as saved with sleuth_save().
CAVEAT: GO terms are not independent, but instead overlap in various ways through their dependencies. Thus, assumptions of the p-value aggregation are not valid. A way to work around this might be the information theoretic approach presented in:
Alterovitz, Gil, Michael Xiang, Mamta Mohan, and Marco F. Ramoni. 2007. “GO PaD: The Gene Ontology Partition Database.” Nucleic Acids Research 35 (suppl_1): D322–27. https://doi.org/10.1093/nar/gkl799.
However, GO PaD does not seem to exist any more, so the partitioning of the GO database would have to be re-implemented.