Evaluation of quantitative structure-activity predictions. Comparison of the predictive power of an artificial intelligence system with human experts

In this paper, we present three new mathematical techniques for evaluating the predictive skills of structure-activity experts. The question addressed in this paper is how to evaluate the predictive ability of structure-activity experts in identifying the most active compounds of a set of drug candidates. The three proposed mathematical techniques are based on the Phi-square Distance, the Rank Comparison, and the Shuffle method, respectively. They have been used to evaluate the performance of a new computer system and three human experts in predicting the antibacterial potencies of a series of chemical compounds in five different biological tests. The expert system, an artificial intelligence structure-activity program called MULTICASE, performed significantly better than one of the human experts and somewhat better than the other two.On leave from the Department of General and Analytical Chemistry, Technical University of Budapest, Szt. Gellert ter 4, 1111 Budapest, Hungary.     

On chip electrofusion of single human B cells and mouse myeloma cells for efficient hybridoma generation

This article describes the development and full characterization of a microfluidic chip for electrofusion of human peripheral blood B-cells and mouse myeloma (NS-1) cells to generate hybridomas. The chip consists of an array of 783 traps, with dimensions that were optimized to obtain a final cell pairing efficiency of 33±6%. B cells were stained with a cytoplasmic stain CFDA to assess the different stages of cell fusion, i.e. dye transfer to NS-1 cells (initiating fusion) and membrane reorganization (advanced fusion). Six DC pulses of 100 μs (2.5 kV/cm) combined with an AC field (30 s, 2 MHz, 500 V/cm) and pronase treatment resulted in the highest electrofusion efficiency of paired cells (51±11%). Hybridoma formation, with a yield of 0.33 and 1.2%, was observed after culturing the fused cells for 14 days in conditioned medium. This work provides valuable leads to improve the current electrofusion protocols for the production of human antibodies for diagnostic and therapeutic applications.     

��There is no such animal (??? ??? ???)����Lessons of a discovery

The discovery of quasicrystals by Dan Shechtman in the early 1980s was a conspicuous event in materials science not only because it led to the production of a plethora of new materials but also because it signified the demise of a dogma in the science of condensed phase materials concerning symmetry restrictions. Having the discovery recognized was not easy and it required stamina on Shechtman??s part. The story of the quasicrystal discovery offers a set of lessons that might be useful to remember in similar situations.     

Multiobjective management of mining under water hazard by game theory

The management of an aquifer is studied under the assumption that the solution of the multiobjective programming model describing the management problem should satisfy a certain set of axioms. It is shown that a certain class of multiobjective problems may be solved by a game-theoretical concept leading to a single objective quasiconvex programming problem. The method is generalization of Nash's cooperative game theoretical model, and may lean on Zeuther's bargaining process. The methodology is applied to the Transdanubian Karstic region in Hungary where three objectives are present: mining costs, water supply and environmental protection. Results are compared with the solution previously obtained by compromise programming with an l₁-norm. It is found that results obtained by the two methods are comparable.     

Experimental strategy for characterization of essential dynamical variables in oscillatory systems: effect of double-layer capacitance on the stability of electrochemical oscillators

An experimentally accessible algorithm for changing the time scale associated with a dynamical variable is proposed. In general, a differential controller can be applied to (a) identify the essential species in oscillatory systems and (b) explore their role in the feedback loops. Here, we report on classifying electrochemical oscillators by changing the time scale over which the electrode potential varies; the type of different electrochemical oscillators is identified based on whether the controlled modification of pseudo-capacitance induces or suppresses current oscillations.     

A mass spectrometric line for tritium analysis of water and noble gas measurements from different water amounts in the range of microlitres and millilitres

This paper describes the procedure followed for noble gas measurements for litres, millilitres and microlitres of water samples in our laboratory, including sample preparation, mass spectrometric measurement procedure, and the complete calibrations. The preparation line extracts dissolved gases from water samples of volumes of 0.2 μ l to 3 l and it separates them as noble and other chemically active gases. Our compact system handles the following measurements: (i) determination of tritium concentration of environmental water samples by the ³He ingrowth method; (ii) noble gas measurements from surface water and groundwater; and (iii) noble gas measurements from fluid inclusions of solid geological archives (e.g. speleothems). As a result, the tritium measurements have a detection limit of 0.012 TU, and the expectation value (between 1 and 20 TU) is within 0.2 % of the real concentrations with a standard deviation of 2.4 %. The reproducibility of noble gas measurements for water samples of 20–40 ml allows us to determine solubility temperatures by an uncertainty better than 0.5 °C. Moreover, noble gas measurements for tiny water amounts (in the microlitre range) show that the results of the performed calibration measurements for most noble gas isotopes occur with a deviation of less than 2 %. Theoretically, these precisions for noble gas concentrations obtained from measurements of waters samples of a few microlitres allow us to determine noble gas temperatures by an uncertainty of less than 1 °C. Here, we present the first noble gas measurements of tiny amounts of artificial water samples prepared under laboratory conditions.     

In celebration of 30 years of Structural Chemistry: launching a series of prefatory reviews

Structural Chemistry -     

Predicting polycyclic aromatic hydrocarbon formation with an automatically generated mechanism for acetylene pyrolysis

Using Reaction Mechanism Generator (RMG), we have automatically constructed a detailed mechanism for acetylene pyrolysis, which predicts formation of polycyclic aromatic hydrocarbons (PAHs) up to pyrene. To improve the data available for formation pathways from naphthalene to pyrene, new high‐pressure limit reaction rate coefficients and species thermochemistry were calculated using a combination of electronic structure data from the literature and new quantum calculations. Pressure‐dependent kinetics for the CH potential energy surface calculated by Zádor et al. were incorporated to ensure accurate pathways for acetylene initiation reactions. After adding these new data into the RMG database, a pressure‐dependent mechanism was generated in a single RMG simulation which captures chemistry from C to C. In general, the RMG‐generated model accurately predicts major species profiles in comparison to plug‐flow reactor data from the literature. The primary shortcoming of the model is that formation of anthracene, phenanthrene, and pyrene are underpredicted, and PAHs beyond pyrene are not captured. Reaction path analysis was performed for the RMG model to identify key pathways. Notable conclusions include the importance of accounting for the acetone impurity in acetylene in accurately predicting formation of odd‐carbon species, the remarkably low contribution of acetylene dimerization to vinylacetylene or diacetylene, and the dominance of the hydrogen abstraction CH addition (HACA) mechanism in the formation pathways to all PAH species in the model. This work demonstrates the improved ability of RMG to model PAH formation, while highlighting the need for more kinetics data for elementary reaction pathways to larger PAHs.