This may be a minor issue, but in circumstances where there is a single gene associated with an annotation under the "annotation_features" key in annotatinos.json, the value associated with each key in the nested dictionary is a string, rather than a list of length 1 with a single string. All other instances are a list of strings. I'm not sure if this is necessary or an intentional design decision but it was giving me some problems when I was looking up associated genes in enrichment results. I think the only reason those annotations were enriched in the first place is because there were very few genes of interest from a very large set of genes.
This may be a minor issue, but in circumstances where there is a single gene associated with an annotation under the "annotation_features" key in annotatinos.json, the value associated with each key in the nested dictionary is a string, rather than a list of length 1 with a single string. All other instances are a list of strings. I'm not sure if this is necessary or an intentional design decision but it was giving me some problems when I was looking up associated genes in enrichment results. I think the only reason those annotations were enriched in the first place is because there were very few genes of interest from a very large set of genes.