2188 Mol. BioSyst., 2012, 8, 2188–2196 This journal is c The Royal Society of Chemistry 2012 Cite this: Mol. BioSyst., 2012, 8, 2188–2196 A ligand-based approach for the in silico discovery of multi-target inhibitors for proteins associated with HIV infectionw Alejandro Speck-Planche,* a Valeria V. Kleandrova, b Feng Luan ac and M. Nata´lia D. S. Cordeiro* a Received 14th March 2012, Accepted 3rd May 2012 DOI: 10.1039/c2mb25093d Acquired immunodeficiency syndrome (AIDS) is a dangerous disease, which damages the immune system cells to the point that the immune system can no longer fight against other infections that it would usually be able to prevent. The causal agent is the human immunodeficiency virus (HIV), and for this reason, the search for more effective chemotherapies against HIV is a challenge for the scientific community. Chemoinformatics and Quantitative Structure–Activity Relationship (QSAR) studies have played an essential role in the design of potent inhibitors for proteins associated with the HIV infection. However, all previous studies took into consideration the discovery of future drug candidates using homogeneous series of compounds against only one protein. This fact limits the use of more efficient anti-HIV chemotherapies. In this work, we develop the first ligand-based approach for the in silico design of multi-target (mt) inhibitors for seven key proteins associated with the HIV infection. Two mt-QSAR models were constructed from a large and heterogeneous database of compounds. The first model was based on linear discriminant analysis (mt-QSAR-LDA) employing fragment-based descriptors. The second model was obtained using artificial neural networks (mt-QSAR-ANN) with global 2D descriptors. Both models correctly classified more than 90% of active and inactive compounds in training and prediction sets. Some fragments were extracted and their contributions to anti-HIV activity through inhibition of the different proteins were calculated using the mt-QSAR-LDA model. New molecules designed from fragments with positive contributions were suggested and correctly predicted by the two models as possible potent and versatile anti-HIV agents. Introduction The human immunodeficiency virus (HIV) is a C-type enveloped retrovirus of the Retroviridae family. 1 This is the causal agent of the acquired immunodeficiency syndrome (AIDS), a severe condition in humans where the immune system begins to fail, allowing other opportunistic infections to occur. 2 HIV/AIDS is recognized by the World Health Organization (WHO) as one of the greatest pandemics. 3 In 2008, a study showed that HIV infections and deaths, since the beginning of the pandemic, exceeded 60 million and 25 million, respectively, with more than 3 million new cases of infected people being estimated per year. 4 Although the treatment of HIV infection consists of highly active antiretroviral therapy (HAART), which was introduced in 1996, 5 until now, HIV/AIDS has been incurable. For this reason, the investigations have been focused on studies of key proteins such as C–C chemokine receptors (CCRs), 6–8 retroviral aspartyl protease (HIV-PR), 9–11 reverse transcrip- tase (HIV-RT), 12–14 and retroviral integrase (HIV-IN). 15–17 Anyway, despite the efforts of the scientific community, the search for new anti-HIV chemotherapies remains as one of the greatest challenges of modern science. In the field of anti-HIV chemotherapies, in silico methodologies have provided essen- tial insights toward the discovery of inhibitors for proteins associated with HIV infection, 18–28 rationalizing the serendipitous organic synthesis in medicinal and pharmaceutical chemistry. 29 However, almost all prior studies have two remarkable disadvantages. From one side, the in silico methodologies employed for the design of anti-HIV agents use small data- bases of analogous compounds. 18–28 On the other hand, the studies were carried out considering only one target like protein. This fact makes impossible the exploration of struc- tural patterns which could be involved in the development of anti-HIV activity. Thus, the discovery of versatile and highly a REQUIMTE/Department of Chemistry and Biochemistry, University of Porto, 4169-007 Porto, Portugal. E-mail: alejspivanovich@yahoo.es, ncordeir@fc.up.pt; Fax: +351 220402659, +351 220402659 b Faculty of Technology and Production Management, Moscow State University of Food Production, Volokolamskoe shosse 11, Moscow, Russia c Department of Applied Chemistry, Yantai University, Yantai 264005, People’s Republic of China w Electronic supplementary information (ESI) available. See DOI: 10.1039/c2mb25093d Molecular BioSystems Dynamic Article Links www.rsc.org/molecularbiosystems PAPER