hydrophobicity and hydrogen bond capacities, allowing them to
interact with many parts on the protein surface. A fast Fourier
transform correlation approach is implemented to sample millions
of receptor–probe complexes. These complexes are clustered within
the same probe type, and the generated clusters are ranked based on
their binding energies. The method extracts top-ranking clusters
from different types of probe and regroups them again. The con-
sensus sites (CSs) represent putative hot spots, where multiple
probes congregate. The CSs containing the largest number of
probe clusters are considered the main hot spots (Fig. 3)
[31]. The FTMAP method has been developed as web server,
which is linked to the protein data bank (PDB) and generates a
file containing the CSs and their congregated probes for a given
target [34].2.2 Druggability
Evaluation Methods
Once a binding site is identified, the next stage focuses on the
evaluation of its druggability. In general, this is usually accom-
plished by analyzing the binding site using various geometrical
descriptors and binding free energy analysis. All these calculations
are summarized in what is termed as a druggability index. A drugg-
ability index,ID, provides a measure of the potential of a given
binding site to accommodate a small-molecule candidate drug.Fig. 3Probe-based binding site identification mechanism. Different types of drug-like fragments dock
randomly on the protein surface. The first clusters are ranked based on their binding free energy. After the
second turn of clustering, multiple types of fragments are gathering around the consensus sites (CS)
Prediction of Druggable Binding Sites 91