Annual Report 2004

Return to archive

The problem of a classification of the E.coli promoters with respect to their electrostatic potentials was studied at LIT in cooperation with the Institute of Theoretical and Experimental Biophysics and the Institute of Cell Research of RAS (Pushchino). The classification of promoters and other functionally important genome fragments according to their nucleotide sequences and physical-chemical properties is a key factor for understanding gene transcription, replication, recombination and their regulation. This work presents the approach that allows computation of electrostatic potentials of long nucleotide sequences of DNA for both procaryotic and eucaryotic species. The electrostatic potentials of E.coli promoters and periodic sequences were calculated. The electrostatic characteristics of the genome promoters, together with the primary structure, are expected to provide their reliable classification [25].

The results of recent studies on the development of new statistical models of stock market data were presented. On the basis of the analysis of a large number of stocks of various companies, it is shown that for some stock market data the statistical distribution of closing prices normalized by corresponding traded volumes (the index called by the authors as "Price/Volume ratio") fits well a log-normal law. For most stocks such a correspondence is reached with no additional detrending procedure. For other stocks, the distribution has a more complicated character and in most cases is described by a weighed sum of some functions of the log-normal distribution. However, after application of a detrending procedure all considered data can be described by a single log-normal distribution [26].

The problem of robust extraction of trend and chaotic components from stock market time series was considered. The proposed methods also allow one to extract a part of the chaotic component, the so-called anomalous term, which is caused by the transient short-time surges with high amplitudes. This provides more accurate determination of the trend component. The methods are based on the M-evaluation with decision functions of Huber and Tukey type. The iterative numerical schemes for determination of trend and chaotic components are briefly presented, which is resulting in a desired solution within a finite number of iterations. The optimal level for extraction of the chaotic component is determined by a new numerical scheme based on the fractal dimension of the chaotic component of the analyzed series. The forecasting scheme that uses the realized part of the analyzed series and a priori expert information was discussed [27].

© Laboratory of Information Technologies, JINR, Dubna, 2005