Article Preview
TopIntroduction
Human immunodeficiency virus, or HIV, is the virus that causes acquired immune deficiency syndrome (AIDS). The HIV attacks the immune system by damaging or killing a specific type of white blood cell called a T-lymphocyte, also called a CD4+ or T-helper cell (Conner & Villarreal, 2013; Moss, 2013; Murphy & Weaver, 2016; Nowak & McMechael,1995).
Non-nucleoside reverse transcriptase inhibitors (NNRTIs), are a class of antiretroviral drugs (Cihlar & Ray, 2010; Zhan et al., 2013; Li et al., 2012; King et al., 2002), bind to and block HIV reverse transcriptase (an HIV enzyme). HIV uses polymerase also known as reverse transcriptase (RT) to convert its RNA into DNA . The enzyme synthesizes a double stranded DNA copy of the genomic RNA. The DNA molecule can then integrate into the host genome and become a functional component of its genetic makeup (Temin & Baltimore, 1972). These inhibitors include TIBO, HEPT, Nevirapine, Pyridinone, BHAP, and R-APA. Among them HEPT has proved to be a potent and selective inhibitor of HIV-1(Garg et al., 1999).
1-[(2-hydroxyethoxy) methyl] -6- (phenyl thio) thymine, known as the HETP derivatives are an important series of non-nucleoside inhibitors described in 1989 (Miyasaka et al., 1989) and was revealed as being a potent and selective inhibitor of HIV-1, HETP is one of the most selective drugs due to its high specificity and less toxicity (Baba et al.,1989).
Anti-HIV inhibition QSAR models have been developed (Luco & Ferretti, 1997; Douali et al., 2004; Shaik et al., 2013; Hansch & Zhang, 1992; Inthajak et al., 2017; Prakasvudhisarn & Lawtrakul, 2008; Arakawa et al., 2006; Hannonghua et al., 1996; Toropova et al., 2014; Weekes & Fogel, 2003) for HEPT derivatives (Figure 1). Using different training sets these models are based on Multiple Linear Regression (MLR) (Luco & Ferretti, 1997), Partial Least Squares (PLS) (Luco & Ferretti, 1997), Artificial Neural Network (ANN) (Douali et al., 2004) and Support Vector Machines (SVM) (Shaik et al., 2013). Several structural descriptors and physicochemical variables were used to characterize the studied HEPT derivatives.
Table 1. Statistical parameters of different constructed QSAR models
| Statistical methods | R2 | Number of molecules | Ref |
| ANN | 0.958
0.825 | 80
107 | (Douali et al., 2004)
(Shaik et al., 2013 |
| MLR | 0.799
0.944
0.951 | 107
33
107 | (Shaik et al., 2013)
(Hansch & Zhang, 1992)
(Luco & Ferretti, 1997) |
| PLS | 0.943
0.944 | 79
107 | (Luco & Ferretti, 1997)
(Luco & Ferretti, 1997) |
| SVM | 0.817 | 107 | (Shaik et al., 2013) |