The present paper focuses on determining the number of PLS components by using resampling methods such as cross validation (CV), Monte Carlo cross validation (MCCV), bootstrapping (BS), etc. To resample the training data, random non-negative weights are assigned to the original training samples and a sample-weighted PLS model is developed without increasing the computational burden much. Random weighting is a generalization of the traditional resampling methods and is expected to have a lower risk of getting an insufficient training set. For prediction, only the training samples with random weights less than a threshold value are selected to ensure that the prediction samples have less influence on training. For complicated data, because the optimal number of PLS components is often not unique or readily distinguished and there might exist an optimal region of model complexity, the distribution of prediction errors can be more useful than a single value of root mean squared error of prediction (RMSEP). Therefore, the distribution of prediction errors are estimated by repeated random sample weighting and used to determine model complexity. RSW is compared with its traditional counterparts like CV, MCCV, BS and a recently proposed randomization test method to demonstrate its usefulness. Copyright © 2010 John Wiley & Sons, Ltd.

A highly selective fluorescence sensor was developed for Hg(II) ion detection in aqueous solutions based on the selective binding of Hg(II) ions with a pair of thymine-thymine mismatch. The sensor consists of two DNA probes functionalized with a fluorophore (fluorescein, F) and a quencher (tetramethyl rhodamine, Q) moiety separately. This pair of DNA probes contains two pairs of thymine-thymine mismatches used to detect Hg(II) ions. In the presence of Hg(II) ions, thymine-Hg²⁺-thymine was formed between thymine residues of probes. From that, the interaction of the two DNA probes increased. Thus, the DNA probes formed a double-stranded structure. Both the fluorophore and quencher were brought close to each other leading to the fluorescence resonance energy transfer (FRET) between F and Q. Under the optimum conditions, the sensor was used to detect the Hg(II) ions from 50 to 1000 nmol·L⁻¹ with a regression equation y=5281.13−1650.56 lg[Hg²⁺] (R²=0.985). The linear range covers 100 to 500 nmol·L⁻¹, and the limit of detection (LOD) is 79 nmol·L⁻¹. The disturbance of some co-existing metal ions was explored, and no significant fluorescence quenching in the presence of 1.0 μmol·L⁻¹ other metal ions was observed. The fluorescence sensor has good sensitivity and selectivity for Hg(II) ions providing a rapid, simple and low cost method for the detection of mercury(II) ions in aqueous solutions.

A simple and rapid colorimetric approach for the determination of adenosine has been developed via target inducing aptamer structure switching, thus leading to Au colloidal solution aggregation. In the absence of the analytes, the aptamer/gold nanoparticle (Au NP) solution remained well dispersed under a given high ionic strength condition in that the random-coil aptamer was readily wrapped on the surface of the Au NPs, which resulted in the enhancement of the repulsive force between the nanoparticles due to the high negative charge density of DNA molecules. While in the presence of adenosine, target-aptamer complexes were formed and the conformation of the aptamer was changed to a folded structure which disfavored its adsorption on the Au NP surface, thus leading to the reduction of the negative charge density on each Au NP and then the reduced degree of electrostatic repulsion between Au nanoparticles. As a result, the aggregation of the Au colloidal solution occurred. The changes of the absorption spectrum could be easily monitored by a UV-Vis spectrophotometer. A linear correlation exists between the ratio of the absorbance of the system at 522 to 700 nm (A_{522 nm}/A_{700 nm}) and the concentration of adenosine between 100 nmol·L⁻¹ and 10 µmol·L⁻¹, with a detection limit of 51.5 nmol·L⁻¹.

By simply adding ascorbic acid in advance of AgNO₃, the size and shape controllable Au/Ag bimetallic nanoparticles (NP) were prepared in the traditional Au growth solution free of seed at room temperature. The size distribution of NP is well uniform with ca. 10%–15% standard deviation in diameter. By changing CTAB concentration, the size and shape of NPs are tunable. After researching the surface-enhanced Raman spectroscopy (SERS) behavior of the prepared NPs, an enhancement factor varied from 4.3×10⁴ to 1.1×10⁵ was obtained for the NP centered at ca. (64±8) nm. Electrochemical cyclic voltammetric results revealed that the so formed nanoparticles were Au riched Au/Ag bimetallic NP, and this formation might be due to the disproportionation reaction of Au⁺ prompted by Ag⁺ and the under potential deposition process of Ag⁺ on Au.