RESUMEN

Lignocellulosic biomass is mainly composed of cellulose, hemicellulose, and lignin. Fuzzy logic, in turn, is a branch of many-valued logic based on the paradigm of inference under vagueness. This paper presents a methodology, based on computational intelligence, for modeling the kinetics of a complex reactional system. The design of a fuzzy interpolator to model cellulose hydrolysis is reported, within the perspective of applying kinetic models in bioreactor engineering. Experimental data for various types of lignocellulosic materials were used to develop the interpolator. New experimental data from the enzymatic hydrolysis of a synthetic substrate, on the other hand, were used to validate the methodology. The accuracy of the results indicates that this is a promising approach to extend the application of models fitted for specific situations to different cases, thus enhancing their generality.

Asunto(s)

Lignina/metabolismo , Modelos Químicos , Animales , Reactores Biológicos , Celulasa/metabolismo , Análisis por Conglomerados , Hidrólisis , Cinética , Trichoderma/enzimología

RESUMEN

Bacillus megaterium is a promising host for expression of heterologous proteins. This paper reports the nutrient consumption patterns and production of metabolites for three different strains of B. megaterium, ATCC 14945, QMB 1551 and PV 361, which is QMB 1551 with seven constitutive plasmids deleted. 14 h cultivations in agitated flasks were run, for two different media: A (LB plus 10g/L glucose) and B (medium A, with the yeast extract replaced by tryptone). Strains PV361 and QMB 1551 showed higher maximum specific growth rates in medium B, reaching 0.42 h-1 and 0.48 h-1 respectively. The main by-products of the glucose overflow mechanism were acetate and lactate, for all three strains, which had preferential amino acids for substrate: Ala, Asp, Glu, Ser. No production of alcohols was detected.