Abstract
Schizophrenia is a long-term, serious mental health condition that affects how a person thinks, perceives, and behaves. This disorder often results in substantial difficulties in social interactions and work performance. Individuals with schizophrenia might appear disconnected from reality, causing significant distress both for themselves and their Friends. Although symptoms of schizophrenia can vary from person to person, they typically fall into three main categories: cognitive, negative, and psychotic. Creating computational tools to find and develop drugs for schizophrenia has more interest in the past few years. Regardless of the significant developments in drug design, the fundamental approach still uses topological descriptors. Topological indices are used to estimate the bioactivity of chemical compounds in QSAR/QSPR studies. In general, with the use of the quantitative structure-property relationship (QSPR), topological indices are numerical values that are connected to the chemical drug structures and are used to predict their reactivity, stability, and properties. This work focuses on calculating different eccentric indices (EIs), developing a regression model for thirteen anti-schizophrenia drugs, and applying statistical methods to establish a linear regression relationship between QSPR correlating properties and eccentric indices. Statistical analysis shows that p-values less than equals 0.05, f-test value ( > 2.5) , and values of correlation r are greater than 0.7 validate the calculations. The correlation coefficient (r2) is a convenient tool for evaluating the QSPR models' quality. r2 > 0.7 is essential for a good QSPR model. The p-values show the significance of the results, while the correlation coefficient values show the accuracy of the results. In order to fit regression models for the calculated eccentric index values, eight physicochemical properties of anti-schizophrenia drugs are examined. Drug properties like molar refractivity (cm(3)), refractive index (cm(3)), enthalpy (kJ/mol), melting, boiling and flash points (°C), complexity, and molecular weight are all more effectively estimated by the QSPR model. By examining actual and estimated values for the drugs, the results are verified.