Table 1.

Data availability per resource and organism (human, mouse, rat and pig)

Resource/annotation	Human	Mouse	Rat	Pig
Ensembl genome annotations
Coding genes	20 418	22 619	22 250	22 452
Non-coding genes	22 107	15 795	8 934	3 250
eggNOG mammalian orthology
Coding genes assigned to orthologous groups	86.7%	84.3%	76.4%	82.2%
Mentions in biomedical literature
Organism	–	1 824 080	1 629 280	133 937
Gene	–	1 304 170	734 243	57 230
Gene Ontology annotations
Experimental	107 301	89 360	49 281	817
Author statement	48 894	4 760	3 396	27
Inferred	86 785	170 033	188 718	47 225
Electronic	74 049	44 022	40 559	101 074
High-scoring STRING protein-protein interactions
Experimental	18 069	1 304	920	1 266
Experimental transferred	12 713	22 030	39 381	32 312
TISSUES expression data
Experimental datasets	4	4	3	3
Tissues covered by experimental data	20	20	12	20

The number of coding and non-coding genes for the assemblies of human, mouse, rat, and pig in Ensembl release 95 are reported. From the eggNOG v4.5 orthology database, we report the percentage of genes from each organism that are assigned to a mammalian orthologous group. Text mining of all PubMed abstracts and a subset of full text articles available from PMC provided the number of publications that mention each organism and its genes. We grouped the most recent Gene Ontology annotations into four categories based on their evidence codes and counted the number of annotations for each group in each organism. High-scoring protein–protein interactions from the STRING v10.5 database (overall confidence score above 0.7) were counted. For the TISSUES 2.0 database of mammalian expression, the number of experimental datasets supporting the 21 main tissues is reported together with the number of tissues covered by these datasets. See Supplementary Table S1 as well as Methods for more details.