The subject of invention is the asphalt with the addition of PMMA / ATH composite powder, which replaces the filler or is a supplement to the filler. PMMA / ATH composite powder is added according to the invention to the filler to improve properties of asphalt mixtures. The invention falls within the scope of asphalt. The filler is the smallest sand fraction in the asphalt mixture with standardized sieving curve. Asphalt mixture according to the invention contains in weight percentage from 3% to 8% of binder, from 75% to 94% of stone aggregate with a particle size greater than the grain size of filler, from 0 % to 17% of filler, and from 0.5 to 17% of PMMA / ATH composite powder.
F.07 Improvements to an existing product
COBISS.SI-ID: 4621850Organisation of a workshop for European project BioChemLig (Bio-Orthogonal Chemo-Specific Ligation) We prepared the learning materials: NOVIČ, Marjana (editor). BioChemLig Meeting : 8. - 11. July 2012, Ljubljana. [Ljubljana]: [Kemijski inštitut], 2012. 2 zv. (58; 99), ilustr. [COBISS.SI-ID 5016602] Six invited lecturers and two researchers delivered lectures and exercises for 20 participants from EU countries (France, Swiss, England, Italy, Slovenia).
B.01 Organiser of a scientific meeting
COBISS.SI-ID: 5016602In silico approach for identification of potent antioxidants acting as free radical scavengers is presented, which is an alternative approach that can reduce expensive and time consuming in vivo or in vitro studies. Free radical formation in living cells is connected to a number of diseases as well as to aging due to high reactivity causing cell damage that accumulates over time. The consequence of irreparable DNA damage is apoptosis; nonlethal DNA mutations may cause the formation of a cancer, or replication and transcriptional errors triggering viral infections. The predictive computational models may replace expensive, time consuming and ethically questionable experiments [1]. In this study we present extended models for prediction of antioxidants potency of structurally diverse compounds, elaborated from our previous work [2]. The nonlinear modeling method is used to specify the relationship between chemical structures and antioxidative potency. The counter-propagation neural network (CP ANN) model performed with the RMS cross-validation error below 14%. Previously detected activity of coumarine derivatives has been further elaborated by including new experimental data on these compounds into the training set. The new prediction models were developed and compared for different modeling methods, CP ANN, Multiple Linear Regression and Support Vector Regression. They show good performance and may serve as a prioritization scheme for further synthesis of new antioxidants.
B.04 Guest lecture
COBISS.SI-ID: 5189402Indo-US Workshop on Mathematical Chemistry has been organized since 1995 in India or USA and is aimed for students of chemistry and related areas. In the lecture we presented basics of the QSAR modelling. This includes: the collecting of data and organizing of data bases, calculation of molecular descriptors, application of different modelling methods and testing and validation of models. The main areas of QSAR modeling are the drug discovery (for searching of new and better drugs), the toxicology and eco-toxicology. We presented some QSAR models, which are available over internet, and their application on the set of conazoles and halocarbons.
B.04 Guest lecture
COBISS.SI-ID: 5189658A short history of graphical approach to bioinformatics is presented. It was initiated in 1983 by Hamory, by representing DNA sequence as a path in 3D space. Another important contribution was of Jeffrey, who in 1990 modified the Chaos Game for representation of DNA. In 2000, Randić initiated matrix representation of DNA graphical curves and maps, which contributed to quantitative (numerical) representation of bio-sequences. This approach transformed visual analysis of DNA to mathematical analysis. In following years several additional graphical representations were introduced, including the spectrum-like representations, which have led to graphical approach to the alignment of DNA and proteins. Most recently Randić reported an approach based on suitable modifications of amino-acid adjacency matrix. This allows solving the problem of protein alignment rigorously without use of trial-and-error methodology, statistical methods and empirical parameters, as do the current methods.
B.04 Guest lecture
COBISS.SI-ID: 5002522