Line retrieves enriched GO BP terms computed with a standard and a network-based procedure. Both are performed with Bonferroni-corrected Fisher tests, considering a significance level of 5 . We benchmarked on the OMIM-derived benchmark set the level of annotation added by the network-based method from both a quantitative and qualitative point of view. The quantitative analysis highlights the ability of the networkbased method in recovering new enriched functions. The qualitative analysis focuses on six cases for which the newly (S)-1-Boc-2-Hydroxymethyl-piperazine enriched Cyclohept-2-enone terms add new biological insights, as confirmed by previously published experimental data.Quantitative analysis on OMIM diseasesFor assessing the power of the network-based enrichment, we focus uniquely on GO BP terms that are not enriched by the standard method (filtering out also all the terms that are ancestors of terms enriched by the standard method). Results are listed in Table 1. In eleven cases out of 244 (5 ), neither the standard enrichment nor the network-based enrichment retrieve significantly overrepresented BP term (first PubMed ID:https://www.ncbi.nlm.nih.gov/pubmed/17184506 row in Table 1). In 143 cases (58 ) the network-based enrichment detects more terms than the standard one (last two rows in Table 1). The average number of these terms is 38 per disease.Moreover, in 86 cases (35 ) the network-based procedure is able to enrich terms that were not included in the sets of annotations characterizing the input protein set (last row in Table 1). The average number of these new terms is 17. It is also worth noticing that the networkbased enrichment returns significant terms in 7 cases out of the 18 where the PubMed ID:https://www.ncbi.nlm.nih.gov/pubmed/9084065 standard method fails to provide any result (data not shown). 30 of the annotations refer to GO terms that are associated to less than 100 proteins in the human proteome, describing quite specific functions. Terms that are more common are less frequently enriched, mainly owing to the Bonferroni-corrected Fisher test that we applied (see Figure 3). Network-based methods introduce a bias towards terms associated to the most connected nodes (see in our case Figure 1S, Additional file 2). We find that the bias is also present in the case of the standard enrichment procedure that does not make use of the network information (Figure 1S).Qualitative analysis on OMIM diseasesThe newly enriched terms that are absent in the original annotations of the input genes are likely to gain new knowledge on the disease at hand. We focus the qualitative analysis on them and we detail here six case studies for which experimental validations are available for the annotations derived with our method. For all the reportedDi Lena et al. BMC Genomics 2015, 4-dione 4-Bromo-3-hydroxypyridine Dibenzo[b 16(Suppl 8):S6 http://www.biomedcentral.com/1471-2164/16/S8/SPage 7 ofTable 1. Functional annotation of 244 OMIM diseases with our pipelineAnnotation* No significant GO BP terms extracted by SE and NET-GE Same significant terms extracted by SE and NET-GE NET-GE enriches more terms already included in the annotation of the input proteins NET-GE adds new terms not included in the annotation of the input proteins OMIM diseases?11 (5 ) 90 (37 ) 57 (23 ) 86 (35 )*Functional annotation is performed with our network-based procedure (NET-GE) and with a standard enrichment (SE) method. ?number out of the 244 OMIM diseases.cases, PINA does not return any significant association. EnrichNet enriches only terms that are already included in the annotations of the input proteins. However EnrichNet is best suited to analyze sets i.