Figure 1.
A sub-graph of the DAG for DO term ‘Cutaneous lupus erythematosus (DOID:0050169)’.
The arrow symbol represents an ‘IS_A’ link of DO. For example, “Cutaneous lupus erythematosus (DOID:0050169)” is linked to “Skin disease (DOID:37)” by an ‘IS_A’ relationship.
Table 1.
Data sources used for measuring disease similarity.
Figure 2.
d1, d2 are two diseases, and dMICA is the MICA of d1 and d2. G1, G2 and GMICA represent gene sets related to d1, d2 and dMICA, respectively.
Figure 3.
A. The similarities of disease pairs from the benchmark set and 100 random sets were calculated by SemFunSim, FunSim, Resnik, Wang, BOG, and PSB. B. The similarities of all the disease pairs between 916 diseases with PTCs in CTD were measured by SemFunSim, FunSim, Resnik, Wang, BOG, and PSB. In addition, the similarities of all the disease pairs between these 916 diseases with PTCs and 44 diseases without PTCs in CTD were computed by SemFunSim.
Figure 4.
AUC analysis of the benchmark set and random sets.
A. ROC curves for the experimental results on the benchmark set and a random set. It shows 1-specificity versus sensitivity of each method for calculating the similarities of disease pairs. B. Average of AUC for 100 permutations.
Figure 5.
The number of pairs of diseases identified using SemFunSim with common PTCs.
A. The number of pairs of the T100-PDDs with common PTCs. B. The number of pairs of the T100-PSDs with common PTCs. The yellow area represents the number of pairs without common PTCs. The pink area indicates the number of pairs with common PTCs and adjusted P-value ≥0.05. The light blue area represents the number of pairs with common PTCs and adjusted P-value <0.05.
Figure 6.
The number of pairs of similar diseases identified using the five methods with common PTCs.
Blue bar indicates the number of pairs with common PTCs. Red bar represents the number of pairs with common PTCs and adjusted P-value <0.05.
Figure 7.
The number of pairs of similar diseases identified using MimMiner and SemFunSim with common PTCs.
A. The number of pairs of the top pairs of similar diseases with common PTCs. The red bar represents the number of pairs with common PTCs measured by SemFunSim. The blue bar indicates the number of pairs with common PTCs measured by MimMiner. B. The number of pairs of the top pairs of similar diseases with common PTCs and adjusted P-value <0.05. The red bar represents the number of pairs with common PTCs and adjusted P-value <0.05 measured by SemFunSim. The blue bar indicates the number of pairs with common PTCs and adjusted P-value <0.05 measured by MimMiner.
Table 2.
Top 20 pairs of similar diseases.
Table 3.
Associations between PTCs and diseases retrieved from PubMed.