Search

In this paper, we propose a novel clustering algorithm which uses a weighted combination of several criteria as its fitness function. We demonstrate the suitability of the new method in the case of clustering text documents. The proposed algorithm leverages the concept of nearest neighbour separation (NNS) to enhance the separation of the clusters and also outlines a heuristic to compute the NNS. A new parameterized fitness function has been proposed which can be tuned to provide more weightage to the traditional metrics based on inter- and intra-cluster distances of clusters or on the NNS. Genetic Algorithm has been used to perform the actual clustering and the results obtained has been compared with the traditional K-Means algorithm. The performance of the algorithm has been tested on different standard datasets, and the results have been presented.

First-derivative electron paramagnetic resonance (EPR) absorption spectra of the vanadyl ion (VO²⁺) in glasses of five different organic−aqueous cosolvent mixtures at 10 K are reported over the 9.4−376 GHz microwave frequency range. Whereas absorption features arising from the g-parallel and g-perpendicular components of the ge tensor are overlapping at low frequencies, it is shown that they are completely separated at microwave frequencies >110 GHz. The analysis of the EPR spectra revealed that line-broadening is due to g-strain over the 9.4−376 GHz range. EPR spectra of VO²⁺ bound to bovine transferrin in a 1:1 molar ratio were also compared at 9.4 and 110 GHz. The results for the VO²⁺−transferrin complex showed that absorption features corresponding to the g-parallel and g-perpendicular components of the ge tensor were similarly separated at high frequency and that line-broadening was attributable to g-strain. We show that line-broadening characteristic of g-strain under conditions of low g-anisotropy can be explained by an S = ¹/2 spin system in which the principal g values are themselves the random variables described by a normal distribution. On this basis, it is shown that line-broadening is proportional to the microwave frequency and the spread of each of the principal values of the ge tensor.

Nephrocalcin (NC) is a calcium-binding glycoprotein of 14,000 molecular weight. It inhibits the growth of calcium oxalate monohydrate crystals in renal tubules. The NC used in this study was isolated from bovine kidney tissue and purified with the use of DEAE-cellulose chromatography into four isoforms, designated as fractions A-D. They differ primarily according to the content of phosphate and gamma-carboxy-glutamic acid. Fractions A and B are strong inhibitors of the growth of calcium oxalate monohydrate crystal, whereas fractions C and D inhibit crystal growth weakly. Fraction A, with the highest Ca(2+)-binding affinity, was characterized with respect to its metal-binding sites by using the vanadyl ion (VO2+) as a paramagnetic probe in electron paramagnetic resonance (EPR) and electron nuclear double resonance (ENDOR) spectroscopic studies. By EPR spectrometric titration, it was shown that fraction A of NC bound VO2+ with a stoichiometry of metal:protein binding of 4:1. Also, the binding of VO2+ to NC was shown to be competitive with Ca2+. Only protein residues were detected by proton ENDOR as ligands, and these ligands bound with complete exclusion of solvent from the inner coordination sphere of the metal ion. This type of metal-binding environment, as derived from VO(2+)-reconstituted NC, differs significantly from the binding sites in other Ca(2+)-binding proteins.

(MeOH and H2O/MeOH mixtures).