Skip to main content

Table 4 Effectiveness of clustering of MDDR dataset using QPI: 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 43.84 47.38 48.72 50.70 53.41 54.06
Cosine 45.60 46.08 47.56 50.46 53.79 54.50
Euclidean 44.43 45.54 47.95 48.65 52.68 54.86
Hamming 53.13 56.08 59.07 60.58 64.02 67.76
Jaccard 57.86 60.62 64.07 66.49 70.68 73.53
Manhattan 56.01 58.10 60.99 61.86 64.56 65.97
CSPA Correlation 46.81 50.04 51.72 51.78 54.23 56.36
Cosine 46.04 49.49 51.42 52.11 54.48 55.92
Euclidean 46.20 49.86 51.05 51.88 54.36 56.33
Hamming 54.67 58.50 60.27 61.78 62.33 65.66
Jaccard 55.03 59.13 60.84 61.03 63.73 67.44
Manhattan 55.08 59.00 59.10 60.84 61.78 64.61
HGPA Correlation 47.59 49.51 52.39 54.45 56.86 58.56
Cosine 45.58 48.44 52.78 54.42 56.36 58.70
Euclidean 46.92 51.41 53.20 54.75 57.00 58.97
Hamming 55.24 58.48 60.30 63.99 68.21 69.22
Jaccard 55.71 59.89 64.10 65.15 70.48 71.60
Manhattan 54.84 58.98 62.73 63.58 65.85 69.97
Individual clustering Ward's method   52.33 54.86 56.90 59.00 61.33 63.17
\