Skip to main content

Table 2 Effectiveness of clustering of MDDR dataset using F-Measure: ALOGP Fingerprint

From: Voting-based consensus clustering for combining multiple clusterings of chemical structures

Clustering method No. of clusters
500 600 700 800 900 1000
Consensus clustering CVAA Correlation 26.80 21.96 18.96 18.49 17.6 15.45
Cosine 24.79 21.72 19.01 18.19 16.46 14.81
Euclidean 27.96 23.75 22.68 24.30 21.17 19.95
Hamming 24.02 20.48 16.31 16.85 14.95 14.68
Jaccard 23.58 21.96 18.01 18.46 16.72 15.35
Manhattan 27.03 25.23 21.16 20.36 19.10 19.05
CSPA Correlation 5.06 4.65 4.16 3.56 3.35 3.04
Cosine 5.17 4.65 4.08 3.62 3.37 3.05
Euclidean 5.12 4.64 4.04 3.61 3.38 3.00
Hamming 5.30 4.74 4.16 3.62 3.54 3.13
Jaccard 5.31 4.82 4.15 3.77 3.48 3.13
Manhattan 5.33 4.80 4.21 3.62 3.45 3.05
HGPA Correlation 7.13 5.48 5.45 4.65 4.35 4.37
Cosine 8.06 6.04 5.03 4.52 4.45 4.08
Euclidean 7.08 6.55 5.65 4.67 4.56 4.60
Hamming 8.37 5.73 4.94 5.29 4.97 4.93
Jaccard 7.63 6.22 5.98 4.53 5.24 3.92
Manhattan 7.72 6.48 5.23 5.35 4.90 4.12
Individual clustering Ward's method   9.93 9.19 8.19 7.17 6.67 6.44
\