20 sets of synthesized molecular compounds were considered as anti-breast cancer agent. The molecular descriptors which describe the microbial activities of the studied compounds were calculated using B3LYP/6-31+G* and obtained. The parameters obtained were EHOMO (eV), ELUMO (eV), dipole moment (Debye), log P, molecular weight (amu), HBA, HBD, Vol and Ovality were screened. The obtained calculated descriptors were used to develop QSAR model for prediction of experimental inhibition concentration (IC50) using SPSS and Gretl software packages for multiple linear regression (MLR) and MATLAB for artificial neural network (ANN). From this statistical analysis, MLR and ANN were observed to be predictive, however, ANN-QSAR model predicted more efficiently than MLR. Furthermore, molecular docking study was executed with breast cancer cell line (PDB ID: 1hi7); it was observed that BS20 with binding energy of -7.0 kcal/mol bounded more efficiently than other compounds also, it inhibited more than the standard used (5-FU).