Quantitative structure-activity relationship studies of TIBO derivatives using support vector machines

Abstract

A quantitative structure-activity relationship (QSAR) study is suggested for the prediction of anti-HIV activity of tetrahydroimidazo[4,5,1-jk][1,4]benzodiazepinone (TIBO) derivatives. The model was produced by using the support vector machine (SVM) technique to develop quantitative relationships between the anti-HIV activity and ten molecular descriptors of 89 TIBO derivatives. The performance and predictive capability of the SVM method were investigated and compared with other techniques such as artificial neural networks and multiple linear regression. The results obtained indicate that the SVM model with the kernel radial basis function can be successfully used to predict the anti-HIV activity of TIBO derivatives with only ten molecular descriptors that can be calculated directly from only molecular structure. The contribution of each descriptor to the structure-activity relationships was evaluated. Hydrophobicity of the molecule was thus found to take the most relevant part in the molecular description.

HIVprotI: an integrated web based platform for prediction and design of HIV proteins inhibitors

“…Researchers have put a considerable focus on HIV therapy and a lot of compounds have been tested against this pathogen [ 3 , 4 ]. However, a few antiretroviral drugs have been able to slow the disease progression.…”

In Silico SAR Studies of HIV-1 Inhibitors

“…The tetrahydroimidazo[4,5,1-jk][1,4]benzodiazepines (TIBOs), as non-nucleoside analogues, constitute a group of potent inhibitors of HIV-1 RT. This family of compounds was the subject of several 2D and 3D QSAR studies [ 4 , 5 , 6 , 7 , 8 ], and many models have been developed from this database. Research work carried out on this family of compounds treated only correlation problems.…”

Εφαρμογές Πολυπαραμετρικών Στατιστικών Τεχνικών Στη Χημική Ανάλυση

“…Ο Zhang et al [49] δίνει ένα εύρος για τον αριθµό των µονάδων στις ενδιά- + a, όπου a θετικός αριθµός που κυµαίνεται από 1 ως 10 καθορίζοντας τον ελάχιστο και µέγιστο αριθµό των νευρώνων. Ο Darnag et al [50] αναφέρεται στην παράµετρο ρ = n t / w n και περιορίζει την τιµή της στο διάστηµα: 1 (ή 0,9) < ρ < 2,2 (ή 3). Αν ρ << 1, το µοντέλο αποµνηµονεύει τα δείγµατα εκπαίδευσης, ενώ αν ρ >> 3, το µοντέλο δεν µπορεί να γενικεύσει.…”

“…Οµάδα Ι (Class I): αναφέρεται σε δίκτυα τα οποία έχουν περισσότερες εισερχόµενες µεταβλητές από εξερχόµενες. Τέλος, η Gramatica [58] προτείνει και ο Darnag et al [50] χρησιµοποιεί πιο προηγµένους δείκτες για την εύρεση του AD του βέλτιστου µοντέλου. Συγκεκριµένα, ελέγχεται η "ισχύς" ("leverage") ενός άγνωστου δείγµατος σε σχέση µε την οµάδα εκπαίδευσης.…”