Table 2 Prediction precision by evidence code and NAF threshold with H. sapiens as the target.

Pair	Measure	NAF	#pred	PIP	#val	Precision
Evidence code EXP
SC-HS	EC	2	383	1903	159	41.5%
SC-HS	EC	4	73	1903	30	41.1%
SC-HS	S³	2	399	1903	150	37.6%
SC-HS	S³	4	68	1903	22	32.4%
SC-HS	Importance	2	412	1903	125	30.3%
SP-HS	EC	2	49	1906	22	44.9%
SP-HS	EC	4	7	1906	3	42.9%
Evidence code IPI
SC-HS	ec+s3+Imp	4	10,917	19278	2305	21.1%
SC-HS	ec+s3+Imp	7	19,075	19,278	6137	32.2%
SC-HS	ec+s3+Imp	11	12,341	19,278	5436	44.0%
SC-HS	ec+s3+Imp	17	8568	19,278	4655	54.3%
SC-HS	ec+s3+Imp	29	2961	19,278	1838	62.1%
AT-HS	ec+s3+Imp	2	10,600	7740	3471	32.7%
AT-HS	ec+s3+Imp	4	6757	7740	3304	48.9%
AT-HS	ec+s3+Imp	7	3416	7740	1945	56.9%
AT-HS	ec+s3+Imp	11	1564	7740	937	59.9%
AT-HS	ec+s3+Imp	17	610	7740	393	64.4%
AT-HS	ec+s3+Imp	29	109	7740	79	72.5%
CE-HS	ec+s3+Imp	2	8652	3703	4200	48.5%
CE-HS	ec+s3+Imp	4	4482	3703	2603	58.1%
CE-HS	ec+s3+Imp	7	1998	3703	1253	62.7%
CE-HS	ec+s3+Imp	11	637	3703	410	64.4%
CE-HS	ec+s3+Imp	17	133	3703	88	66.2%
CE-HS	ec+s3+Imp	29	15	3703	10	66.7%
DM-HS	ec+s3+Imp	2	18,858	8439	5701	30.2%
DM-HS	ec+s3+Imp	4	10,119	8439	5473	54.1%
DM-HS	ec+s3+Imp	7	6310	8439	3566	56.5%
DM-HS	ec+s3+Imp	11	7546	8439	4358	57.8%
DM-HS	ec+s3+Imp	17	3360	8439	2025	60.3%
DM-HS	ec+s3+Imp	29	59	8439	43	72.9%
SP-HS	ec+s3+Imp	2	8244	6974	2755	33.4%
SP-HS	ec+s3+Imp	4	3026	6974	1669	55.2%
SP-HS	ec+s3+Imp	7	1677	6974	1054	62.9%
SP-HS	ec+s3+Imp	11	786	6974	529	67.3%
SP-HS	ec+s3+Imp	17	347	6974	235	67.7%
SP-HS	ec+s3+Imp	29	71	6974	53	74.6%
MM-HS	ec+s3+Imp	2	8469	13394	3191	37.7%
MM-HS	ec+s3+Imp	4	2629	13394	1489	56.6%
MM-HS	ec+s3+Imp	7	849	13394	544	64.1%
MM-HS	ec+s3+Imp	11	239	13394	148	61.9%
MM-HS	ec+s3+Imp	17	31	13394	13	41.9%
MM-HS	ec+s3+Imp	29	8	13394	5	62.5%

This table summarizes prediction precision as a function of NAF for species aligned with human satisfying the degree-3 criterion. The species pairs are presented in order of mean CCS degree, highest to lowest (cf. Supplementary Table 1). We show predictions based on source evidence codes EXP (top section) and IPI (bottom section) available in 2010 and validated (with any evidence) in 2020. PIP means predictable in principle, and refers to the absolute maximum number of predictions that would be possible in principle given the information available as of April 2010 (see text). To save space in the IPI case, we have conglomerated the measures EC, S³, and Importance, since all three had similar validation rates at fixed NAF (within 10% of each other in all cases). Also to save space, not all values of NAF are listed here, but the Pearson correlation between NAF and precision across all NAF values are presented below, in Table 3.

Search