Background: The quantitative structure-activity relationship is an analysis method that
can be applied for designing new molecules. In 1997, Hopfinger and coworkers developed the 4DQSAR
methodology aiming to eliminate the question of which conformation to use in a QSAR
study. In this work, the 4D-QSAR methodology was used to quantitatively determine the influence
of structural descriptors on the activity of aryl pyrimidine derivatives as inhibitors of the TGF-β1
receptor. The members of the TGF-β subfamily are interesting molecular targets, since they play
an important function in the growth and development of cell cellular including proliferation, apoptosis,
differentiation, Epithelial-Mesenchymal Transition (EMT), and migration. In late stages,
TGF-β exerts tumor-promoting effects, increasing tumor invasiveness, and metastasis. Therefore,
TGF-β is an attractive target for cancer therapy.
Objective: The major goal of the current research is to develop 4D-QSAR models aiming to propose
new structures of aryl pyrimidine derivatives.
Materials and Methods: Molecular dynamics simulation was carried out to generate the conformational
ensemble profile of a data set with aryl pyrimidine derivatives. The conformations were overlaid
into a three-dimensional cubic box, according to the three-ordered atom alignment. The occupation
of the grid cells by the interaction of pharmacophore elements provides the Grid Cell Occupancy
Descriptors (GCOD), the dependent variables used to build the 4D-QSAR models. The best
models were validated (internal and external validation) using several statistical parameters. Docking
molecular studies were performed to better understand the binding mode of pyrimidine derivatives
inside the TGF-β active site.
Results: The 4D-QSAR model presented seven descriptors and acceptable statistical parameters
(R2 = 0.89, q2 = 0.68, R2
pred = 0.65, r2
m = 0.55, R2
P = 0.68 and R2
rand = 0.21) besides pharmacophores
groups important for the activity of these compounds. The molecular docking studies helped to understand
the pharmacophoric groups and proposed substituents that increase the potency of aryl
Conclusion: The best QSAR model showed adequate statistical parameters that ensure their fitness,
robustness, and predictivity. Structural modifications were assessed, and five new structures
were proposed as candidates for a drug for cancer treatment.