Biological-Activity Predictions, Crystallographic Comparison and Role of Packing Interactions in Androstane Derivatives of Steroids
- First Online:
- Cite this article as:
- Chand, B. & Malik, M.A. J Chem Crystallogr (2011) 41: 255. doi:10.1007/s10870-010-9979-0
- 64 Downloads
A total of 60 molecules of androstane derivatives of steroids (1–60) have been undertaken to predict their pharmacological effects, specific mechanisms of action, known toxicities, drug likeness, etc., by using the statistics of multilevel neighbourhoods of atoms (MNA) descriptors for active and inactive fragments. The biological activity spectra for substances have been correlated on SAR base (Structure–activity relationships data and knowledge base) which provides the different Pa (probability of activity) and Pi (probability of inactivity). The Lipinski’s rule predicts that all the androstane derivatives have stronger preponderance for “cancer-like-drug” molecules and some of their related analogous have been entered in the ANCI (American National Cancer Institute) database. Some selected bond distances and bond angles of interest have been taken into account and deviation of bond distances/bond angles, vis-a-vis the substitutional group and X–H···A intra/intermolecular hydrogen bonds have been discussed in detail. X–H···A intra/intermolecular hydrogen bonds in the identified molecules have been described with the standard distance and angle cut-off criteria. D–θ and d–θ scatter plots for X–H···A intra-and intermolecular interactions are presented for better understanding of packing interactions existing among these derivatives. Comparison of contacts from H(C) to O and H(O) to O, vis-a-vis their crystal structure reveals that contacts from H(O) to O predominate over H(C) to O. Solvent–solute/solute–solvent interactions have also been investigated to understand more complicated processes that occur for biomolecules in aqueous solutions. Most of the molecules show high value of drug-likeness whereas molecule-3 (82.5%), 36 (87.2%), 41 (83.7%), 43 (86.5%) and 50 (85.9%) exhibit low value of drug-likeness, instead of observed range of 90.3–99.2%.