Naslov
Computing ecotoxicology - actuality and perspectives

Autori
Hackenberger Kutuzović, Branimir

Vrsta, podvrsta i kategorija rada
Sažeci sa skupova, sažetak, znanstveni

Izvornik
Abstracts of the 5th Croatian Congress of toxicology with Interantional Participation ; u: Archives of Industrial Hygiene and Toxicology 2016 ; 67 (Suppl. 1) / - , 2016

Skup
5th Croatian Congress of Toxicology with International Participation

Mjesto i datum
Poreč, Hrvatska, 09.10.2016. - 12.10.2016

Vrsta sudjelovanja
Pozvano predavanje

Vrsta recenzije
Međunarodna recenzija

Ključne riječi
DEFENSoil ; ecotoxicological experiments ; GPU and CUDA software ; predictive toxicology

Sažetak
Computational technology has been one of crucial drivers of science progress over the past half-century. Today, the development is unimaginable without disciplines such as bioinformatics, computational chemistry, computational biology, etc. Computers have ceased to be a tool whose main purpose is to accelerate calculation or store a large amount of data. Now they play a role in decision making and adaptation of algorithms based on the power of parallel computing methods and the opportunities of deep learning, neural networks, and artifiial intelligence methods. Forward scientifi march of toxicology has resulted in the blossoming of new areas ripe for further investigation. The major part of new disciplines such as –omics, high-throughput screening and, particularly, nanotoxicology rely on the creation of new computational capabilities and a new software. In addition to a signifiant role in predictive toxicology (mainly assessment of toxic effects of untested chemicals), computational methods are increasingly being used in ecotoxicology. They enable computational experiments which have a strong correlation with realistic conditions, often impossible to obtain in practice. Within DEFENSoil project, methods for computational toxicological and ecotoxicological experiments are being developed and tested against conventional toxicological tests. The preliminary results show a high degree of correlation at all experimental levels. The use of hardware based on GPU and CUDA based software enables the development of highly realistic simulations in the fist phase of experiment and execution of experiment in the second phase.

Izvorni jezik
Engleski

Znanstvena područja
Biologija

POVEZANOST RADA