Search

Multivariate saccharide quantitative structure-activity relationships (QSARs) have been developed for two series of 9-hydroxyellipticine glycosides. In order to describe the structural variation of the glycoside moieties, thirteen chromatographic variables were used. Eleven D-glycosides and seven L-glycosides were used in separate QSARs to model 9-hydroxyellipticine glycoside antitumour activity against L 1210 leukemia. The multivariate partial least squares (PLS) method was used to establish the QSARs.

The concentrations of 17 trace elements (e.g., copper, cobalt, iron, manganese, chromium, silicon and magnesium) were determined in whole blood samples of 81 persons working with different welding methods on stainless steel or aluminium and 68 nonwelders. Inorganic spark source mass spectrometry was used for the chemical analyses. The data were analyzed by the SIMCA method for pattern recognition (discriminant analysis). No differences were found between the five groups, either in the average levels of the trace elements or in the correlation structures between the trace elements. Thus no blood concentration data on the analyzed elements and collected from a single person contained any information with respect to exposure to the welding fumes investigated.

The SIMCA method of pattern recognition (PaRC) was used to analyze structure-activity data for a series of phenethylamine agonists and antagonists of the beta-adrenergic receptor. On the basis of physicochemical substituent parameters the SIMCA method classified correctly 100% of the agonists and 88% of the antagonists. In addition, parameters derived from the class models were correlated with the biological activities of the agonists and antagonists, respectively. Test compounds not included in the initial data analysis were classified and their activities estimated. The applicability of pattern recognition in structure-activity studies in general is discussed.

The effects of thioridazine and thiothixene were studied by a double-blind technique on 40 schizophrenic patients. The doses were adjusted for optimal clinical and therapeutic effects and side effects were rated after three and eight weeks of treatment. No statistically significant differences were observed between the two drugs or between either of the two drugs and the previous medication. Plasma levels were estimated by a fluorometric technique after three and eight weeks of treatment. No correlation was found between plasma levels and clinical effects for either thioridazine or thiothixene. Plasma levels of both drugs were clearly correlated to dosage after three weeks of treatment. After eight weeks this correlation persisted for thioridazine but not for thiothixene. By that time plasma levels of thiothixene had decreased to about 30 per cent of the initial value, indicating strong enzyme induction.