The secondary interaction contribution in complexation via multiple hydrogen bonds was estimated using model complexes. Based on the results of hydrogen fluoride complexes, proton donor (D) - proton acceptor (A) type secondary interaction was attractive, and the total stabilization energy should be increased about 25-35%, based on hydrogen bond interaction, by the D-A type secondary interaction. The D-D type secondary interaction was very small. The A-A type secondary interaction was repulsive, and the total stabilization energy should be decreased about 10-20% based on hydrogen bond interaction, by the A-A type secondary interaction.

View full abstract

We have developed a peak identification system for topological fragment spectra (TFS) to support data analysis and knowledge discovery in chemical data mining. A chemical structure can be submitted to the system using a incorporated chemical structure editor. The system can generate the TFS of the molecule and display it on the screen automatically. The peak of interest can be specified on the screen by using a pointing device. Then, structural fragments corresponding to the peak are displayed. The system can also generate synthetic spectra for a series of chemical structures, and is able to analyze their peaks as well. The details of the system and algorithms are described with illustrative examples.

View full abstract

A data set of 420 pesticide persistences in the environment was collected as field half-life (HL) using several on-line databases. Due to the fuzziness of observed values, the compounds were grouped into three major categories: class 1 when a pesticide has HL ≤ 30 days; class 2, if 30 < HL ≤ 100 days; and class 3, if HL > 100 days. The Quantitative Structure-Biodegradation Relationship (QSBR) analysis was worked out on the training set of 315 pesticides. Thirty one topological substructural descriptors were used and the decision tree approach was employed for the modeling. Estimation results were as follows: for the train set, 5 compounds of two-unity (class 1/class 3) misclassification, 38 compounds of unity (class 1/class 2 and class 2/class 3) misclassification, and 272 compounds (86.3%) were correctly classified; for the test set, there were 3, 20, and 82 compounds (78.1%) respectively. The computer expert system EKeeper was developed on the basis of the QSBR model.

View full abstract

An open-source electrochemical measurement system called ECHAN (ElectroCHemical ANalyzer) has been developed with Visual Basic Language. This software enables automated measurement of electrode samples with macro language successively through various electrochemical methods. Cyclic voltammetry, rotating ring-disk voltammetry, potential step chronoamperometry, current vs. rotation speed measurement and multipotentiostat measurement are all carried out with this single software. It is also equipped with a searching function by keywords and measurement conditions. After measurement, data can be processed by FFT (Fast Fourier Transform) low pass filter, smoothing by the simple moving average or the Savitzky-Golay method and operations between two data sets. It makes possible the evaluation of electro-catalysis by the Koutecky-Levich analysis, the Levich analysis, the electron number concerned with oxygen reduction reaction, chronoamperometry analysis by Cottrell's equation, coulomb number calculation and the Tafel analysis. Since an A/D board is used for data sampling, this program is applicable to a conventional potentiostat, which is usually connected to an X-Y recorder.

View full abstract

In the domain of chemistry, many experimental data are being accumulated by combinatorial chemistry and high throughput screening (HTS) and so on. In order to obtain benefit information from these data, a lot of methodologies of chemometrics have been developed. Since usually these numerical data are represented as multi-dimensional arrays, traditional statistical software cannot handle these data directly. In this study, we developed data modeling software, which handles multi-dimensional arrays effectively. By using this software we are able to analyze multi-dimensional data. It is possible to understand the structure of multi-dimensional data by using graphical representation such as bar graph, line graph and so on. And the data modeling method, Multi-way PLS (partial least squares) and Kohonen neural network are available for the analysis of multi- dimensional data.

View full abstract