Challenges in the construction of knowledge bases for human microbiome-disease associations

Table 1 Existing datasets for disease- and species-related entities. Note that there are only two datasets which contain both diseases and species (miRNA and variome). In addition, species-level datasets are not specific to the human microbiome, so there is a need to create datasets curated for human microbiota

Dataset	Entity type	No. of annotations	No. of unique annotations
CDR [34]	Disease	12,694	3459
Variome [35]	Disease	6025	629
miRNA [36]	Disease	2123	671
NCBI Disease [37]	Disease	6881	2129
Arizona Disease [38]	Disease	3206	1188
SCAI [39]	Disease	2226	1048
CellFinder [40]	Species	435	51
Variome [35]	Species	182	8
miRNA [36]	Species	726	47
S800 [41]	Species	3646	1564
LocText [42]	Species	276	39
Linneaus [43]	Species	4077	419
BioNLP-ST 16 [43]	Species	619	277