Nearest Neighbor Classification Rule for Mixtures of Discrete and Continuous Random Variables

In this paper very simple nonparametric classification rule for mixtures of discrete and continuous random variables is described. It is based on the method of nearest neighbor proposed by Cover and Hart (1967). The bounds on the limit of the nearest neighbor rule risks are given. Both lower and upper bound depend on the Bayes risk and the loss function. Finally the method is compared with other existing methods on some practical data set.     

On the choice of the optimal periodic operation for a continuous fermentation process

In this contribution we investigate the impact of the forcing waveform on the productivity of a continuous bioreactor governed by an unstructured, nonlinear kinetic model. The (periodic) forcing is applied on the substrate concentration in the feed. To this end, some alternative waveforms commonly encountered in practice are evaluated and their performance is compared. An analytical/numerical approach is used. The preliminary analytical step is based on the π‐criterion that gives useful information for small amplitudes. The extension to larger amplitudes, when significant improvements are expected, is then performed through a continuation‐optimization procedure. It is found that the choice of the specific waveform has an impact on the performance of the process and there is no unique best forcing for any process condition, but its choice depends on the operating parameters and the forcing amplitude and frequency values. Further, the influence of the waveform functions on the wash‐out conditions are extensively examined. The analysis shows that all the waveforms examined in this work may lead to significant enlargement of the nontrivial regime with respect to a steady state operation. In particular, square‐wave forcing leads in practice to the extinction of the wash‐out conditions for any feed substrate concentration and for a well defined choice of the forcing parameters. © 2010 American Institute of Chemical Engineers Biotechnol. Prog., 2010     

The membrane dialysis bioreactor with integrated radial-flow fixed bed —a new approach for continuous cultivation of animal cells

A hybridoma cell was cultivated continuously in a membrane dialysis bioreactor with an integrated radial-flow fixed bed consisting of porous Siran^® carriers over a period of 6 weeks. Antibodies accumulated to an average of 100 mg l^?1, approx. 10 times more than in fixed bed cultures without dialysis membrane. Serum costs could be reduced about 85% due to an appropriate feeding strategy. Siran^® carriers with 3–5 mm diameter showed an advantage compared to those with 1–2 mm diameter. For the 3–5 mm carrier the specific glucose uptake rate and the MAb production rate were constant, if the velocity was between 0.09 mm s^?1 and 0.75 mm s^?1. At higher velocities cells are washed out of the bed. Furthermore antibody consistency and cell stability were verified in long-term cultivations over a period of 96 days. From an estimation of the antibody concentration reachable with the reactor concept under optimal conditions a concentration 45 times higher compared to axial-flow fixed bed reactors and 11 times higher compared to stirred tank reactors can be expected.     

The use of periphyton communities for nutrient removal from polluted streams

The results of experiments on the efficiency of periphyton communities for nutrient removal from polluted streams in a continuous flow-through are given. The artifical stream (5 m × 0.7 m × 0.5 m) was made of wood, with silon (a kind of nylon) screens, as a substratum for periphyton growth. The elimination of nutrients was monitored by ammonium, nitrite, nitrate and orthophosphate analyses. In addition, the elimination of organics and the decrease in trophic state were determined. During two field experiments a marked elimination of nitrogen and phosphorus was demonstrated. The maximum efficiency of ammonium and orthophosphate removal was 80% and 70%, respectively. Organic removal reached 35% (C.O.D._Mn-Kubel) and 54% (B.O.D.₅). Inorganic and organic nutrient elimination caused significant changes in periphyton community structure in the outflow portion of the through, evaluated by the saprobic index and the similarity coefficient. The experiments confirmed that periphyton communities are a useful means of nutrient removal from polluted streams.     

Growth and survival ofAzospirillum brasilense andArthrobacter giacomelloi in binary continuous culture

Summary Azospirillum brasilense andArthrobacter giacomelloi were grown in single and mixed succinate-limited continuous cultures at a partial oxygen pressure of 0.01atm. Growth, viability and survival during nutrient starvation were examined at various dilution rates. At D=0.05 h^–1, K_s values for succinate consumed were calculated.Arthrobacter giacomelloi viability was inversely related to dilution rate whereasAzo. brasilense was directly related. Slightly lower values of viability were obtained in mixed culture, but the ratio between the microorganisms was constant. The survival ofArth. giacomelloi in single culture decreased with increasing growth rate while survival ofAzo. brasilense was directly related to dilution rate. Acetylene reduction activity was generally very low in both single and mixed cultures. Respiration rate was also determined and the mixed culture showed an oxygen uptake rate higher than that of single cultures.Research work supported by CNR, Italy. Special grant I.P.R.A. Sub-project 1. Paper N. 317.     

Metabolic,physiological and kinetic aspects of the alcoholic fermentation of whey permeate by Kluyveromyces fragilis NRRL 665 and Kluyveromyces lactis NCYC 571

Optimum growth conditions for the fermentation of non-concentrated whey permeate by Kluyveromyces fragilis NRRL 665 have been defined. Use of 3.75 g yeast extract l^?1, a growth temperature of 38°C and a pH of 4.0 allowed a maximum productivity of 5.23 g ethanol l^?1 h^?1 in continuous culture with a yield 91% of theoretical. Complete batch fermentation of permeate with 100 g lactose l^?1 was possible with a maximum specific growth rate of 0.276 h^?1 without any change in ethanol yield. Fermentation of concentrated permeate resulted, however, in a general decrease of specific substrate consumption rate, demonstrated by the inability to completely convert an initial 90 or 150 g lactose l^?1 in continuous culture, even at dilution rates as low as 0.05 and 0.08 h^?1, respectively. The decrease could be related to substrate inhibition, to an increase in osmotic pressure caused by lactose and salts, and to ethanol inhibition of both alcohol and biomass yield. The decrease in specific productivity could be counterbalanced by use of high cell density cultures, obtained by cell recycle of K. fragilis. Fermentation of a non-concentrated permeáte at a dilution rate of 1 h^?1 resulted in a productivity of 22 g l^?1 h^?1 at 22 g ethanol l^?1. Cell recycle using flocculating Kluyveromyces lactis NCYC 571 was also tested. With this strain a productivity of 9.3 g l^?1 h^?1 at 45 g product l^?1 was attained at a dilution rate of 0.2 h^?1, with an initial lactose concentration of 95 g l^?1.     

Anaerobic degradation of 3,4,5-trimethoxybenzoate by a defined mixed culture of Acetobacterium woodii, Pelobacter acidigallici, and Desulfobacter postgatei

Abstract Acetobacterium woodii was continuously grown on 3,4,5-trimethoxybenzoate as pure culture or in commensalistic combination with Pelobacter acidigallici and Desulfobacter postgatei . Under pure culture conditions the following growth parameters were determined: μ _max= 0.112 h⁻¹, K _s= 1.07 mM, Y _max= 35 g/mol, and m = 0.22 mmol·g⁻¹·h⁻¹. In coculture with P. acidigallici the affinity for the substrate increased and the K _s value was found to be 135 μM. Under batch culture conditions mixed populations of A. woodii, P. acidigallici , and D. postgatei completely mineralized 3,4,5-trimethoxybenzoate to CO₂, whereas under continuous culture conditions more than 3 mM acetate remained unused.     

Toxoplasma gondii: Microassay to differentiate toxoplasma inhibiting factor and interleukin 2

Using a sensitive, economical, and reproducible microassay, the relationship of toxoplasma inhibiting factor to interleukin 2 has been examined. The assay developed took advantage of the observation that (1) Toxoplasma gondii tachyzoites replicated efficiently in the murine monocytic cell line, RAW 264; (2) treatment of RAW 264 cells with toxoplasma inhibiting factor prevented intracellular replication of the parasite to an extent similar to that observed with identical treatment of freshly isolated murine peritoneal exudate cells; and (3) [³H]uracil incorporation was an efficacious means to quantify replication (or inhibition of replication) of tachyzoites within the cell line. Although toxoplasma inhibiting factor and interleukin 2 were both present in the same lectin -and antigen-stimulated splenocyte supernatant fluids, results from microassays strongly suggested that the molecules were two distinct entities.     

Growth kinetics of a bacteriophage in continuous culture

Lytic coliphage Qbeta was grown in continuously cultured host bacteria using a cascade of stirred flow reactors. The apparatus was constructed so that the steady stream of exponentially growing bacterial cells passing through the stirred flow reactors served to prevent coevolution brought about by host-parasite interactions. Wall growth was the primary cause for deviation from ideal continuous culture conditions and is largely dependent on the surface structure of the host bacteria. Using an Escherichia coli strain deficient in adhesive type I pili expression, the desynchronization of single burst events could easily be followed over the course of four infection latency periods. Computer simulations based on a two-stage model for the Qbeta infection cycle were in perfect agreement with the experimental data. Applications of the optimized system to strategies of molecular evolution are discussed. (c) 1996 John Wiley & Sons, Inc.     

Molecular integrity of monoclonal antibodies produced by hybridoma cells in batch culture and in continuous-flow culture with integrated product recovery

The molecular integrity of monoclonal antibodies (MCAB) produced by murine hybridoma cell line TB/C3 was studied in batch and continuous-flow cultures. In batch culture, one band of MCAB was detected initially by Western blotting of sodium dodecyl sulfate (SDS)-polyacrylamide gels run under unreduced conditions, but heterogenous MCAB bands appeared as the culture aged. The latter were due to the degradation of MCAB by proteases active at the neutral pH of the culture. The deleterious effect of proteases was minimized in the continuous-flow cultures which were integrated for product recovery. The MCAB of high quality was purified over 26 days from a culture grown at a dilution rate of 0.025 h(-1) (experiment 1). However, at a lower dilution rate of 0.015 h(-1) (experiment 2), the integrity of MCAB was compromised after the initial 13 days of culture. This was shown to be due to the variation in the carbohydrate content of MCAB produced, as judged by the increased sialylation of heavy chains and the varied reactivity of MCAB with lectins (Maackia amurensis agglutinin, Galanthus nivalis agglutinin, and Datura stramonium agglutinin) as the age of the culture increased. The concentration of the purified MCAB samples by enzyme-linked immunosorbent assay (ELISA) (used normally) was usually higher than that estimated by absorbance at 280 nm. Best correlation between the two methods (ELISA-280 nm ratio of 1.02-1.25) was obtained with experiment 1 samples. This ratio increased in experiment 2 and batch culture samples as the heterogeneity of MCAB produced increased, being 1.03-2.94 and 2.53-4.62, respectively. Therefore, ELISA overestimated MCAB concentration when the molecular integrity of the latter was compromised. The ELISA-A(280) nm ratio might hence provide a useful indicator for assessing the quality of MCAB produced. Comparison of SDS-polyacrylamide gels stained with Coomassie Brilliant Blue R and silver showed that the former correlated better with the MCAB activity stain, whereas the silver stained both the protein- and carbohydrate-rich components. Comparison of the patterns produced with these two stains might therefore offer another parameter to monitor the overall integrity of MCAB produced. Finally, the data presented have important implications on the validity of using long-term and intensive cultures for generating MCAB because such cultures would be subjected to the additive effects reported for batch and continuous modes of growth. (c) 1993 John Wiley & Sons, Inc.