Reconstituting protein interaction networks using parameter-dependent domain-domain interactions

BMC Bioinformatics

Table 3 Yeast protein-domain annotation data after merging annotations from the six databases

Domain annotation set	N _U	N _P		N _S		A _D
Domain annotation set	N _U	n	%	n	%	A _D
Domain-merging procedure
SET-1	2,595	4,709	80.0	1,174,333	40.0	1.05
SET-2	2,847	4,964	84.4	1,510,026	51.7	1.33
SET-3	2,806	5,280	89.7	1,653,122	56.6	1.69
SET-4	2,843	5,307	90.2	1,663,269	56.9	1.69
SET-5	4,182	5,392	91.6	1,735,533	59.4	2.55
SET-6	4,114	5,395	91.7	1,756,481	60.1	2.44
Naïve domain merging
SET-6-NB	10,297	5,395	91.7	1,756,481	60.1	5.77
Domain merging based solely on sequence overlap
SET-6-SB	1,492	5,395	91.7	1,756,481	60.1	1.32

Database sets SET-1 through SET-6 are defined in Table 2. SET-6-NB (naïve merging) contained the union of unique domain annotations from the six databases used in SET-6. SET-6-SB contained merged domain annotations from the same six databases as in SET-6, but domains in this set were merged only if their sequences overlapped and they shared at least ten common amino acids (i.e., domain labels were not considered). N_U, number of unique domains. N_P, proteins with at least one domain annotation. N_S, protein-domain amino acid sequence coverage. A_D, average number of domains per protein.

ISSN: 1471-2105