1. Naiane Sangaletti-Gerharda, Mara Ceab, Vicky Risco, Rodrigo Navia, (2015), In situ biodiesel production from greasy sewage sludge using acid and enzymatic catalysts, Bioresource Technology 179, 63-70.
    [ Show Abstract ]This study proposes to select the most appropriate sewage sludge (greasy, primary and secondary) for in situ transesterification and to compare the technical, economic and energetic performance of an enzymatic catalyst (Novozym®435) with sulfuric acid. Greasy sludge was selected as feedstock for biodiesel production due to its high lipid content (44.4%) and low unsaponifiable matter. Maximum methyl esters yield (61%) was reached when processing the wet sludge using sulfuric acid as catalyst and n-hexane, followed by dried-greasy sludge catalyzed by Novozym®435 (57% methyl esters). Considering the economic point of view, the process using acid catalyst was more favorable compared to Novozym®435 catalyst due to the high cost of lipase. In general, greasy sludge (wet or dried) showed high potential to produce biodiesel. However, further technical adjustments are needed to make biodiesel production by in situ transesterification using acid and enzymatic catalyst feasible. Copyright © 2014 Elsevier Ltd. All rights reserved.

     

  2. E. Chávez, M. Graff, G. Navarro, E. S. Téllez, (2015), Near neighbor searching with K nearest references, Information Systems, 43-61.
    [ Show Abstract ]Proximity searching is the problem of retrieving, from a given database, those objects closest to a query. To avoid exhaustive searching, data structures called indexes are built on the database prior to serving queries. The curse of dimensionality is a well-known problem for indexes: in spaces with sufficiently concentrated distance histograms, no index outperforms an exhaustive scan of the database. In recent years, a number of indexes for approximate proximity searching have been proposed. These are able to cope with the curse of dimensionality in exchange for returning an answer that might be slightly different from the correct one. In this paper we show that many of those recent indexes can be understood as variants of a simple general model based on K-nearest reference signatures. A set of references is chosen from the database, and the signature of each object consists of the K references nearest to the object. At query time, the signature of the query is computed and the search examines only the objects whose signature is close enough to that of the query. Many known and novel indexes are obtained by considering different ways to determine how much detail the signature records (e.g., just the set of nearest references, or also their proximity order to the object, or also their distances to the object, and so on), how the similarity between signatures is defined, and how the parameters are tuned. In addition, we introduce a space-efficient representation for those families of indexes, making it possible to search very large databases in main memory. Small indexes are cache friendly, inducing faster queries. We perform exhaustive experiments comparing several known and new indexes that derive from our framework, evaluating their time performance, memory usage, and quality of approximation. The best indexes outperform the state of the art, offering an attractive balance between all these aspects, and turn out to be excellent choices in many scenarios. Our framework gives high flexibility to design new indexes. KEYWORDS: Proximity search; Searching by content in multimedia databases; k nearest neighbors; Indexing metric spaces

     

  3. Pamela Hidalgo, Gustavo Ciudad, Sigurd Schober, Martin Mittelbach, Rodrigo Navia MG, (2015), Biodiesel synthesis by direct transesterification of microalga Botryococcus braunii with continuous methanol reflux, Bioresource Technology, 181, 32-29.
    [ Show Abstract ]Direct transesterification of Botryococcus braunii with continuous acyl acceptor reflux was evaluated. This method combines in one step lipid extraction and esterification/transesterification. Fatty acid methyl esters (FAME) synthesis by direct conversion of microalgal biomass was carried out using sulfuric acid as catalyst and methanol as acyl acceptor. In this system, once lipids are extracted, they are contacted with the catalyst and methanol reaching 82% wt of FAME yield. To optimize the reaction conditions, a factorial design using surface response methodology was applied. The effects of catalyst concentration and co-solvent concentration were studied. Hexane was used as co-solvent for increasing lipid extraction performance. The incorporation of hexane in the reaction provoked an increase in FAME yield from 82% (pure methanol) to 95% when a 47% v/v of hexane was incorporated in the reaction. However, the selectivity towards non-saponifiable lipids such as sterols was increased, negatively affecting biodiesel quality. KEYWORDS: Botryococcus braunii; Direct transesterification; Acyl acceptor; Co-solvent; Hexane

     

  4. Robinson Muñoz, Rodrigo Navia, Gustavo Ciudad, Catherine Tessini, David Jeison, Rodrigo Mella, Claudia Rabert, Laura Azócar, 2015, Preliminary biorefinery process proposal for protein and biofuels recovery from microalgae, Fuel, 150 425-433.
    [ Show Abstract ]A preliminary study of a new route for the valorization of microalgal biomass including protein and biofuels recovery is presented. The study involved consecutive steps of soluble protein extraction using alkaline conditions, followed by lipid extraction for biodiesel production. After both extraction processes, the spent biomass was used as feedstock for bio-oil production through pyrolysis at 500 °C. The results were compared with pyrolysis using whole microalgal biomass and lignocellulosic feedstock. The new route allowed getting 10% of solubilized protein and only 2% of biodiesel (relative to the total biomass). Bio-oil yield obtained using spent Botryococcus braunii was 33.2%. Higher bio-oil yields were obtained from whole Nannochloropsis gaditana (38.3%), whole B. braunii (39.7%) and pine wood (39.9%). Related to bio-oil characteristics, several protein-derived compounds were identified in bio-oil from spent B. braunii biomass. Therefore, the reduction of these compounds in bio-oil should be a critical target for further research. In addition, considering the low biodiesel production yield, lipid extraction from biomass could be avoided. This would increase lipid derived compounds in bio-oil, improving its heating value. Finally, unlike lignocellulosic biomass derived bio-oil, microalgae biooil showed a neutral pH, preventing possible corrosion problems in combustion engines.

     

  5. Mariela Bustamante, Mario Villarroel, Monica Rubilar, Carolina Shene, (2015) Lactobacillus acidophilus La-05 encapsulated by spray drying: Effect of mucilage and protein from flaxseed (Linum usitatissimum L.), LWT – Food Science and Technology, 62, 1162-1168 .
    [ Show Abstract ]Flaxseed mucilage (FM) and soluble protein (FSP) were used as wall materials for encapsulating Lactobacillus acidophilus La-05 by spray drying. The effects of the content of FM in the encapsulating solution and drying temperature on the survival of bacterium after drying and its viability during storage were evaluated. Optimal conditions for maximizing the survival of L. acidophilus (78%) predicted by the surface response methodology were 0.2% w/v of FM and 110 °C. These encapsulating conditions permitted almost a 2-fold increase of viability compared to the product encapsulated without FM. A further increase in survival (90%) was obtained by supplementing the encapsulating solution with heat-treated FSP. Moreover, the encapsulation of L. acidophilus enhanced its viability during incubation in simulated gastric acid and bile solutions. Keywords Microencapsulation; Spray drying; Lactobacillus acidophilus; Flaxseed mucilage; Flaxseed soluble protein

     

  6. Loreto P. Parra, Giannina Espina, Javier Devia, Oriana Salazar, Barbara Andrews, Juan A. Asenjo, (2015), Identification of lipase encoding genes from Antarctic seawater bacteria using degenerate primers: Expression of a cold-active lipase with high specific activity, Enzyme and Microbial Technology, 68 56-61.
    [ Show Abstract ]Cold-active enzymes are valuable catalysts showing high activity at low and moderate temperatures and low thermostability. Among cold-active enzymes, lipases offer a great potential in detergent, cosmetic, biofuel and food or feed industries. In this paper we describe the identification of novel lipase coding genes and the expression of a lipase with high activity at low temperatures. The genomic DNA from Antarctic seawater bacteria showing lipolytic activity at 4 °C was used to amplify five DNA fragments that partially encode novel lipases using specifically designed COnsensus-DEgenerate Hybrid Oligonucleotide Primers (CODEHOP). All the fragments were found to have a high identity with an α/β-hydrolase domaincontaining protein identified by the sequencing of the complete genome of Shewanella frigidimarina NCIMB 400. The complete sequence of one of the lipasecoding gene fragments, lipE13, was obtained by genome walking. Considering that the other fragments had a high identity to the putative lipase from S. frigidimarina NCIMB 400, the complete lipase genes were amplified using oligonucleotide primers designed based on the 5′ and 3′ regions of the coding sequence of the related protein. This strategy allowed the amplification of 3 lipase-encoding genes of which one was expressed in the periplasm using the Escherichia coli BL21(DE3)/pET-22b(+) expression system. The recombinant protein was obtained with activity toward pnitrophenyl caproate showing a high specific activity between 15 and 25 °C.

     

  7. Miguel A. Campodónico, Barbara A. Andrews, Juan A. Asenjo, Bernhard O. Palsson, Adam M. Feist, (2014), Generation of an atlas for commodity chemical production in Escherichiacoli and a novel pathway prediction algorithm, GEM-Path, Metabolic Engineering, 25, 140-158.
    [ Show Abstract ]The production of 75% of the current drug molecules and 35% of all chemicals could be achieved through bioprocessing (Arundel and Sawaya, 2009). To accelerate the transition from a petroleum-based chemical industry to a sustainable bio-based industry, systems metabolic engineering has emerged to computationally design metabolic pathways for chemical production. Although algorithms able to provide specific metabolic interventions and heterologous production pathways are available, a systematic analysis for all possible production routes to commodity chemicals in Escherichia coli is lacking. Furthermore, a pathway prediction algorithm that combines direct integration of genome-scale models at each step of the search to reduce the search space does not exist. Previous work ( Feist et al., 2010) performed a model-driven evaluation of the growth-coupled production potential for E. coli to produce multiple native compounds from different feedstocks. In this study, we extended this analysis for non-native compounds by using an integrated approach through heterologous pathway integration and growth-coupled metabolite production design. In addition to integration with genome-scale model integration, the GEM-Path algorithm developed in this work also contains a novel approach to address reaction promiscuity. In total, 245 unique synthetic pathways for 20 large volume compounds were predicted. Host metabolism with these synthetic pathways was then analyzed for feasible growth-coupled production and designs could be identified for 1271 of the 6615 conditions evaluated. This study characterizes the potential for E. coli to produce commodity chemicals, and outlines a generic strain design workflow to design production strains. Keywords Systems biology; Pathway predictions; Escherichia coli; Strain design

     

  8. Hidalgo, P., Ciudad, G., Schober S., Mittelbach, M., Navia, R. Improving the FAME yield of in situ transesterification from microalgal biomass through particle size reduction and cosolvent incorporation, Energy and Fuel, 29, 823-822.
    [ Show Abstract ]In situ transesterification of biomass has the potential to simplify and reduce the cost of biodiesel production from microalgae. This route has the advantage of reducing the number of unit operations, by avoiding the lipid extraction step, prior to transesterification reaction. Thus, the biomass is directly contacted with the alcohol and the catalyst required to convert lipids into biodiesel. In this study, essential process parameters for acid in situ transesterification of microalgae biomass to FAME (fatty acid methyl esters) production were evaluated. Parameters such as the particle size of dry microalgae conglomerates and cosolvent use in the reaction mixture were studied, besides of the temperature and catalyst dosage. An experimental Box–Behnken design was used for the optimization of the FAME yield. The best FAME yield obtained under optimized conditions was of 80% wt, using a microalgae particle size fraction of <150 μm, a catalyst dosage of 125% wt, and a reaction temperature of 67 °C and adding 27% vol/vol of petroleum ether as cosolvent in the reaction mixture. According to the obtained results from experimental design, the FAME yield rose with both when the particle size decreased and the cosolvent proportion in the reaction mixture increased. Moreover, adding a cosolvent increased the saturated and monounsaturated methyl ester content in the sample, improving its stability.

     

  9. Rodríguez, V., Lascani, J., Asenjo, J. A., Andrews, B. A., (2015) Production of Cell-Penetrating Peptides in Escherichia coli Using an Intein-Mediated System, Applied Biochemistry and Biotechnology, 175, 3025-3037.
    [ Show Abstract ]Cell-penetrating peptides are molecules with the ability to cross membranes and enter cells. Attention has been put on these peptides as a tool for drug delivery research, as they are able to serve as delivery vectors for large molecules. Intracellular delivery of bioactive peptides is a very promising research area for clinical applications, since peptides are able to simulate protein regions and thus modulate key intracellular protein-protein interactions. Therefore, evaluation of different strategies for production of these peptides is necessary. In this work, an intein-mediated system was used to evaluate Escherichia coli recombinant production of p53pAnt and PNC27 anticancer cell-penetrating peptides. It was demonstrated that the pTXB1 and the pTYB11 vector systems are suitable for production of this kind of peptides. The production process involves a low temperature induction process and an efficient on-column intein-mediated cleavage, which allowed an effective peptide recovery using a single chromatographic step.

     

  10. Guillemin, M. L., Valero, M., Faugeron, S., Nelson, W., Destombe, C., (2014), Tracing the trans-pacific evolutionary history of a domesticated seaweed (Gracilaria chilensis) with archaeological and genetic data, PloS ONE,9, 1-17.
    [ Show Abstract ]The history of a domesticated marine macroalga is studied using archaeological, phylogeographic and population genetic tools. Phylogeographic and population genetic analyses demonstrated that the cultivated red alga Gracilaria chilensis colonised the Chilean coast from New Zealand. Combining archaeological observations with phylogeographic data provided evidence that exchanges between New Zealand and Chile have occurred at least before the Holocene, likely at the end of the Last Glacial Maximum (LGM) and we suggest that migration probably occurred via rafting. Furthermore, the remarkably low microsatellite diversity found in the Chilean populations compared to those in New Zealand is consistent with a recent genetic bottleneck as a result of over-exploitation of natural populations and/or the process of domestication. Therefore, the aquaculture of this seaweed, based essentially on clonal propagation, is occurring from genetically depressed populations and may be driving the species to an extinction vortex in Chile.

     

  11. Rodríguez, V., Asenjo, J. A., Andrews, B. A., (2014), Design and implementation of a high yield production system for recombinant expression of peptides, Microbial Cell Factories,13.
    [ Show Abstract ]Background: Making peptide pharmaceuticals involves challenging processes where many barriers, which include production and manufacture, need to be overcome. A non common but interesting research area is related to peptides with intracellular targets, which opens up new possibilities, allowing the modulation of processes occurring within the cell or interference with signaling pathways. However, if the bioactive sequence requires fusion to a carrier peptide to allow access into the cell, the resulting peptide could be such a length that traditional production could be difficult. The goal of the present study was the development of a flexible recombinant expression and purification system for peptides, as a contribution to the discovery and development of these potentially new drugs. Results: In this work, a high throughput recombinant expression and purification system for production of cell penetrating peptides in Escherichia coli has been designed and implemented. The system designed produces target peptides in an insoluble form by fusion to a hexahistidine tagged ketosteroid isomerase which is then separated by a highly efficient thrombin cleavage reaction procedure. The expression system was tested on the anticancer peptides p53pAnt and PNC27. These peptides comprise the C-terminal region and the N-terminal region of the protein p53, respectively, fused by its carboxyl terminal extreme to the cell penetrating peptide Penetratin. High yields of purified recombinant fused peptides were obtained in both cases; nevertheless, thrombin cleavage reaction was successful only for p53pAnt peptide release. The features of the system, together with the procedure developed, allow achievement of high production yields of over 30 mg of highly pure p53pAnt peptide per g of dry cell mass. It is proposed that the system could be used for production of other peptides at a similar yield. Conclusions: This study provides a system suitable for recombinant production of peptides for scientific research, including biological assays. Keywords: Escherichia coli; Recombinant peptide production; Cell penetrating peptides; Peptide therapeutics; Thrombin cleavage

     

  12. Nam, H., Campodónico, M., Bordbar, A., Hyduke, D.R., Kim, S., Zielinski, D. C., Palsson, B. O., (2014), A Systems Approach to Predict Oncometabolites via Context Specific Genome-Scale Metabolic Networks, PLoS Computational Biology 10.

    [ Show Abstract ]Altered metabolism in cancer cells has been viewed as a passive response required for a malignant transformation. However, this view has changed through the recently described metabolic oncogenic factors: mutated isocitrate dehydrogenases (IDH), succinate dehydrogenase (SDH), and fumarate hydratase (FH) that produce oncometabolites that competitively inhibit epigenetic regulation. In this study, we demonstrate in silico predictions of oncometabolites that have the potential to dysregulate epigenetic controls in nine types of cancer by incorporating massive scale genetic mutation information (collected from more than 1,700 cancer genomes), expression profiling data, and deploying Recon 2 to reconstruct contextspecific genome-scale metabolic models. Our analysis predicted 15 compounds and 24 substructures of potential oncometabolites that could result from the loss-of-function and gain-of-function mutations of metabolic enzymes, respectively. These results suggest a substantial potential for discovering unidentified oncometabolites in various forms of cancers.

     

  13. Merino M. P., Andrews B. A., Asenjo J. A., (2014), Stoichiometric model and flux balance analysis for a mixed culture of Leptospirillum ferriphilum and Ferroplasma acidiphilum, Biotechnol Prog.
    [ Show Abstract ]The oxidation process of sulfide minerals in natural environments is achieved by microbial communities from the Archaea and Bacteria domains. A metabolic reconstruction of two dominant species, Leptospirillum ferriphilum and Ferroplasma acidiphilum, which are always found together as a mixed culture in this natural environments, was made. The metabolic model, composed of 152 internal reactions and 29 transport reactions, describes the main interactions between these species, assuming that both use ferrous iron as energy source, and F. acidiphilum takes advantage of the organic compounds secreted by L. ferriphilum for chemomixotrophic growth. A first metabolic model for a mixed culture used in bacterial leaching is proposed in this article, which pretends to represent the characteristics of the mixed culture in a simplified manner. It was evaluated with experimental data through flux balance analysis (FBA) using as objective function the maximization of biomass. The growth yields on ferrous iron obtained for each microorganism are consistent with experimental data, and the flux distribution obtained allows understanding of the metabolic capabilities of both microorganisms growing together in a bioleaching process. The model was used to simulate the growth of F. acidiphilum on different substrates, to determine in silico which compounds maximize cell growth, and which are essential. Knockout simulations were carried out for L. ferriphilum and F. acidiphilum metabolic models, predicting key enzymes of central metabolism. The results of this analysis are consistent with experimental data from literature, showing a robust behavior of the metabolic model.

     

  14. Gossage L., Pires DE, Olivera-Nappa Á., Asenjo J., Bycroft M., Blundell T. L., Eisen T., An integrated computational approach can classify VHL missense mutations according to risk of clear cell renal carcinoma, Human Molecular Genetics, 23, 5976–5988.
    [ Show Abstract ]Mutations in the von Hippel-Lindau (VHL) gene are pathogenic in VHL disease, congenital polycythaemia and clear cell renal carcinoma (ccRCC). pVHL forms a ternary complex with elongin C and elongin B, critical for pVHL stability and function, which interacts with Cullin-2 and RING-box protein 1 to target hypoxia-inducible factor for polyubiquitination and proteasomal degradation. We describe a comprehensive database of missense VHL mutations linked to experimental and clinical data. We use predictions from in silico tools to link the functional effects of missense VHL mutations to phenotype. The risk of ccRCC in VHL disease is linked to the degree of destabilization resulting from missense mutations. An optimized binary classification system (symphony), which integrates predictions from five in silico methods, can predict the risk of ccRCC associated with VHL missense mutations with high sensitivity and specificity. We use symphony to generate predictions for risk of ccRCC for all possible VHL missense mutations and present these predictions, in association with clinical and experimental data, in a publically available, searchable web server.

     

  15. Diego Arroyuelo, Carolina Bonacic, Veronica Gil-Costa, Mauricio Marín, and Gonzalo Navarro, (2014), Distributed Text Search using Suffix Arrays, Parallel Computing, 40, 471-495.
    [ Show Abstract ]Text search is a classical problem in Computer Science, with many data-intensive applications. For this problem, suffix arrays are among the most widely known and used data structures, enabling fast searches for phrases, terms, substrings and regular expressions in large texts. Potential application domains for these operations include large-scale search services, such as Web search engines, where it is necessary to efficiently process intensive-traffic streams of on-line queries. This paper proposes strategies to enable such services by means of suffix arrays. We introduce techniques for deploying suffix arrays on clusters of distributed-memory processors and then study the processing of multiple queries on the distributed data structure. Even though the cost of individual search operations in sequential (non-distributed) suffix arrays is low in practice, the problem of processing multiple queries on distributed-memory systems, so that hardware resources are used efficiently, is relevant to services aimed at achieving high query throughput at low operational costs. Our theoretical and experimental performance studies show that our proposals are suitable solutions for building efficient and scalable on-line search services based on suffix arrays.

     

  16. Pamela Hidalgo. Gustavo Ciudad, Martin Mittelbach, Rodrigo Navia, (2015), Biodiesel production by direct conversion of Botryococcus braunii lipids: Reaction kinetics modelling and optimization, Fuel, 153, 544–551.
    [ Show Abstract ]Biodiesel production by in situ esterification/transesterification of Botryococcus braunii microalgae biomass was studied in this work. The effect of different experimental parameters including moisture content, catalyst dosage and methanol to co-solvent ratio (v/v) were evaluated in the proposed reaction process. The selected experimental parameters were found to be significant for the in situ reaction process. A theoretical maximum fatty acid methyl esters (FAME) yield of 80.1% was reached using microalgal biomass with a 28.5% moisture content, a catalyst dosage of 148.7 wt%. with respect to total fatty acid content and a methanol:petroleum ether ratio 2.9:1 v/v. The obtained results show that the reaction took place using wet biomass with moisture content up to 30%, due to the high solvent content provoking water dilution in the reaction medium. For reaction kinetics modelling, the experimental data were fit to a simultaneous esterification and transesterification model. The esterification model was used due to the high free fatty acids (FFA) content and suitably represented the experimental data; however transesterification reaction also took place. Thus, a model considering simultaneous esterification and transesterification on overall FAME production rate was proposed. Keywords: In situ transesterification; Microalgae; Reaction kinetics; B. braunii

     

  17. C. Conca, R. Lecaros, J. Ortega, L. Rosier, (2014), Determination of the Calcium channel distribution in the olfactory system, J. Inverse Ill-Posed Problems, 22, 671-711.
    [ Show Abstract ]In this paper we study a linear inverse problem with a biological interpretation, which is modeled by a Fredholm integral equation of the first kind. When the kernel in the Fredholm equation is represented by step functions, we obtain identifiability, stability and reconstruction results. Furthermore, we provide a numerical reconstruction algorithm for the kernel, whose main feature is that a non-regular mesh has to be used to ensure the invertibility of the matrix representing the numerical discretization of the system. Finally, a second identifiability result for a polynomial approximation of degree less than nine of the kernel is also established. Keywords: Inverse problems; olfactory system; kernel determination; partial differential equations; numerical reconstruction

     

  18. Simonet Torres, Rodrigo Navia, Rachel Campbell Murdy, Peter Cooke, Manjusri Misra and Amar K. Mohant, (2015), Green Composites from Residual Microalgae Biomass and Poly(butylene adipate-co-terephthalate): Processing and Plasticization. ACS Sustainable Chemistry & Engineering, 3, 614-62.
    [ Show Abstract ]Innovative biocomposites from residual microalgae biomass (RMB), a byproduct of biodiesel production, and PBAT (poly(butylene adipate-co-terephthalate)) have been prepared in this study. RMB was characterized by Fourier transform infrared spectroscopy (FT-IR) and its thermal stability was determined. Subsequently, RMB and PBAT biocomposites were prepared by extrusion and injection molding. Incorporation of 10, 20 and 30% RMB in the biocomposites was studied. The biocomposites were characterized using FT-IR and thermogravimetric analysis, and their mechanical properties were compared, including tensile, flexural and impact strength. The effect of RMB on the morphology of the polymer matrix was analyzed by scanning electron microscopy and confocal laser scanning microscopy. RMB plasticization was performed with glycerol and urea, comparing different proportions of glycerol and urea. The studies show that it is possible to use RMB in the manufacture of biocomposites with PBAT, obtaining the best extrusion results with 20% RMB. Optimal result was achieved with 30% glycerol and 7.5 phr of urea. Keywords: Biocomposites; microalgae biomass; biodiesel; poly(butylene adipate-coterephthalate); plasticizers; glycerol; urea

     

  19. Diego Arroyuelo, Francisco Claude, Sebastian Maneth, Veli Makinen, Gonzalo Navarro, Kim Nguyen, Jouni Siren, and Niko Valimaki, (2015), Fast in-memory XPath search using compressed indexes Software Practice and Experience, 45, 399–434.
    [ Show Abstract ]Extensible Markup Language (XML) documents consist of text data plus structured data (markup). XPath allows to query both text and structure. Evaluating such hybrid queries is challenging. We present a system for in-memory evaluation of XPath search queries, that is, queries with text and structure predicates, yet without advanced features such as backward axes, arithmetics, and joins. We show that for this query fragment, which contains Forward Core XPath, our system, dubbed Succinct XML Self-Index (‘SXSI’), outperforms existing systems by 1–3 orders of magnitude. SXSI is based on state-of-the-art indexes for text and structure data. It combines two novelties. On one hand, it represents the XML data in a compact indexed form, which allows it to handle larger collections in main memory while supporting powerful search and navigation operations over the text and the structure. On the other hand, it features an execution engine that uses tree automata and cleverly chooses evaluation orders that leverage the speeds of the respective indexes. SXSI is modular and allows seamless replacement of its indexes. This is demonstrated through experiments with (1) a text index specialized for search of bio sequences, and (2) a word-based text index specialized for natural language search.

     

  20. Djamal Belazzougui and Gonzalo Navarro, (2014), Alphabet-independent compressed text indexing, ACM Transactions on Algorithms, 10, 23.
    [ Show Abstract ]Self-indexes can represent a text in asymptotically optimal space under the k-th order entropy model, give access to text substrings, and support indexed pattern searches. Their time complexities are not optimal, however: they always depend on the alphabet size. In this paper we achieve, for the first time, full alphabetindependence in the time complexities of self-indexes, while retaining space optimality. We obtain also some relevant byproducts on compressed suffix trees.

     

  21. Jeremy Barbay, Francisco Claude, Travis Gagie, Gonzalo Navarro, and Yakov Nekrich, (2014), Efficient fully-compressed sequence representations. Algorithmica, 69, 232-268.
    [ Show Abstract ]We present a data structure that stores a sequence s[1..n] over alphabet [1..σ] in nH0(s)+o(n)(H0(s)+1) bits, where H0(s) is the zero-order entropy of s. This structure supports the queries access, rank and select, which are fundamental building blocks for many other compressed data structures, in worst-case time O(lglgσ) and average time O(lgH0(s)). The worst-case complexity matches the best previous results, yet these had been achieved with data structures using nH0(s)+o(nlgσ) bits. On highly compressible sequences the o(nlgσ) bits of the redundancy may be significant compared to the nH0(s) bits that encode the data. Our representation, instead, compresses the redundancy as well. Moreover, our average-case complexity is unprecedented. Our technique is based on partitioning the alphabet into characters of similar frequency. The subsequence corresponding to each group can then be encoded using fast uncompressed representations without harming the overall compression ratios, even in the redundancy. The result also improves upon the best current compressed representations of several other data structures. For example, we achieve (i) compressed redundancy, retaining the best time complexities, for the smallest existing full-text self-indexes; (ii) compressed permutations π with times for π() and π −1() improved to loglogarithmic; and (iii) the first compressed representation of dynamic collections of disjoint sets. We also point out various applications to inverted indexes, suffix arrays, binary relations, and data compressors. Our structure is practical on large alphabets. Our experiments show that, as predicted by theory, it dominates the space/time tradeoff map of all the sequence representations, both in synthetic and application scenarios.

     

  22. Gonzalo Navarro and Sharma Thankachan, (2014), New space/time tradeoffs for top-k document retrieval on sequences. Theoretical, Computer Science, 542 83-97.
    [ Show Abstract ]We address the problem of indexing a collection D={T”1,T”2,…,T”D} of D string documents of total length n, so that we can efficiently answer top-k queries: retrieve k documents most relevant to a pattern P of length p given at query time. There exist linear-space data structures, that is, using O(n) words, that answer such queries in optimal O(p+k) time for an ample set of notions of relevance. However, using linear space is not sufficiently good for large text collections. In this paper we explore how far the space/time tradeoff for this problem can be pushed. We obtain three results: (1) When relevance is measured as term frequency (number of times P appears in a document T”i), an index occupying |CSA|+o(n) bits answers the query in time O(t”s”e”a”r”c”h(p)+klg^2klg^@en), where CSA is a compressed suffix array indexing D, t”s”e”a”r”c”h(p) is its time to find the suffix array interval of P, and @e>0 is any constant. (2) With the same measure of relevance, an index occupying |CSA|+nlgD+o(nlg@s+nlgD) bits answers the query in time O(t”s”e”a”r”c”h(p)+klg^@?k), where lg^@?k is the iterated logarithm of k. (3) When the relevance depends only on the documents, an index occupying |CSA|+O(nlglgn) bits answers the query in O(t”s”e”a”r”c”h(p)+kt”S”A) time, where t”S”A is the time the CSA needs to retrieve a suffix array cell. On our way, we obtain some other results of independent interest.

     

  23. Erika Rosas, Nicolás Hidalgo, Mauricio Marín, Veronica Gil-Costa, (2014), Web Search Results Caching Service for Structured P2P Networks, Future Generation Computer Systems, 30, 254-264.
    [ Show Abstract ]This paper proposes a two-level P2P caching strategy for Web search queries. The design is suitable for a fully distributed service platform based on managed peer boxes (set-top-box or DSL/cable modem) located at the edge of the network, where both boxes and access bandwidth to those boxes are controlled and managed by an ISP provider. Our solution significantly reduces user query traffic going outside of the ISP provider to get query results from the respective Web search engine. Web users are usually very reactive to worldwide events which cause highly dynamic query traffic patterns leading to load imbalance across peers. Our solution contains a strategy to quickly ease imbalance on peers and spread communication flow among participating peers. Each peer maintains a local result cache used to keep the answers for queries originated in the peer itself and queries for which the peer is responsible for by contacting the Web search engine on-demand. When query traffic is predominantly routed to a few responsible peers our strategy replicates the role of “being responsible for” to neighboring peers so that they can absorb query traffic. This is a fairly slow and adaptive process that we call mid-term load balancing. To achieve a short-term fair distribution of queries we introduce a location cache in each peer which keeps pointers to peers that have already requested the same queries in the recent past. This lets these peers share their query answers with newly requesting peers. This process is fast as these popular queries are usually cached in the first DHT hop of a requesting peer which quickly tends to redistribute load among more and more peers. Keywords Web search engines; Caching services; Load balancing; P2P networks

     

  24. Veronica Gil-Costa, Carolina Bonacic, Alonso Inostrosa, Jair Lobos, Mauricio Marín, (2014), Modelling Search Engines Performance using Coloured Petri Nets, Fundamenta Informaticae, 131, 139-166.
    [ Show Abstract ]This paper proposes using Coloured Petri Nets to model performance of vertical search engines for Web search. In such systems, queries submitted by users or client systems are handled by different components implemented as services deployed on large clusters of dedicated processors. We propose models that represent key features of components running time cost at a suitable level of detail. A comprehensive evaluation study is presented to reveal good precision of models when compared against actual implementations and complex process-oriented simulators of the same search engine instances. A C++ class library is proposed to enable rapid model construction by using a hierarchical and scalable approach, and to enable transparent generation and efficient execution of respective simulation programs either sequentially or in parallel.

     

  25. Mauricio Ramírez-Castrillón, Sandra Denise Camargo Mendes, Mario Inostroza Ponta, Patricia Valente, (2014), (GTG)5 MSP-PCR Fingerprinting as a Technique for Discrimination of Wine, Associated Yeasts?, Plos One, 9.
    [ Show Abstract ]In microbiology, identification of all isolates by sequencing is still unfeasible in small research laboratories. Therefore, many yeast diversity studies follow a screening procedure consisting of clustering the yeast isolates using MSP-PCR fingerprinting, followed by identification of one or a few selected representatives of each cluster by sequencing. Although this procedure has been widely applied in the literature, it has not been properly validated. We evaluated a standardized protocol using MSP-PCR fingerprinting with the primers (GTG)5 and M13 for the discrimination of wine associated yeasts in South Brazil. Two datasets were used: yeasts isolated from bottled wines and vineyard environments. We compared the discriminatory power of both primers in a subset of 16 strains, choosing the primer (GTG)5 for further evaluation. Afterwards, we applied this technique to 245 strains, and compared the results with the identification obtained by partial sequencing of the LSU rRNA gene, considered as the gold standard. An array matrix was constructed for each dataset and used as input for clustering with two methods (hierarchical dendrograms and QAPGrid layout). For both yeast datasets, unrelated species were clustered in the same group. The sensitivity score of (GTG)5 MSP-PCR fingerprinting was high, but specificity was low. As a conclusion, the yeast diversity inferred in several previous studies may have been underestimated and some isolates were probably misidentified due to the compliance to this screening procedure.

     

  26. P. Cumsille, J. Asenjo, C. Conca, A novel model for biofilm growth and its resolution by using the hybrid immersed interface-level set method, Computers and Mathematics with Applications 67 34-51 ene-14
    [ Show Abstract ]In this work we propose a new model to simulate biofilm structures (”finger-like”, as well as, compact structures) as a result of microbial growth in different environmental conditions. At the same time, the numerical method that we use in order to carry out the computational simulations is new to the biological community, as far as we know. The use of our model sheds light on the biological process of biofilm formation since it simulates some central issues of biofilm growth: the pattern formation of heterogeneous structures, such as finger-like structures, in a substrate-transportlimited regime, and the formation of more compact structures, in a growth-limitedregime. The main advantage of our approach is that we consider several of the most relevant aspects of biofilm modeling, particularly, the existence and evolution of a biofilm-liquid interface. At the same time, in order to perform numerical simulations, we have used sophisticated numerical techniques based on mixing the immersed interface method and the level-set method, which are well described in the present work.

     

  27. Courtney K. Robinson, Jacek Wierzchos, Celeste Black, Alexander Crits-Christoph, Bing Ma, Jacques Ravel, Carmen Ascaso, Octavio Artieda, Sergio Valea, Mónica Roldán, Benito Gómez-Silva and Jocelyne DiRuggiero. Microbial diversity and the presence of algae in halite endolithic communities are correlated to atmospheric moisture in the hyper-arid zone of the Atacama Desert, Enviromental Microbiology 17 299-315 feb-15
    [ Show Abstract ]The Atacama Desert is one of the oldest and driest deserts in the world, and its hyper-arid core is described as ‘the most barren region imaginable’. We used a combination of high-throughput sequencing and microscopy methods to characterize the endolithic microbial assemblages of halite pinnacles (salt rocks) collected in several hyper-arid areas of the desert. We found communities dominated by archaea that relied on a single phylotype of Halothece cyanobacteria for primary production. A few other phylotypes of salt-adapted bacteria and archaea, including Salinibacter, Halorhabdus, and Halococcus were major components of the halite communities, indicating specific adaptations to the unique halite environments. Multivariate statistical analyses of diversity metrics clearly separated the halite communities from that of the surrounding soil in the Yungay area. These analyses also revealed distribution patterns of halite communities correlated with atmospheric moisture. Microbial endolithic communities from halites exposed to coastal fogs and high relative humidity were more diverse; their archaeal and bacterial assemblages were accompanied by a novel algae related to oceanic picoplankton of the Mamiellales. In contrast, we did not find any algae in the Yungay pinnacles, suggesting that the environmental conditions in this habitat might be too extreme for eukaryotic photosynthetic life.

     

  28. Petr Vítek, Jan Jehlička, Carmen Ascaso, Vlastimil Mašek, Benito Gómez-Silva, Héctor Olivares, Jacek Wierzchos. Distribution of scytonemin in endolithic microbial communities from halite crusts in the hyperarid zone of the Atacama Desert, Chile. FEMS Microbiology Ecology 90 351-366 nov-14
    [ Show Abstract ]Scytonemin, a UV-screening molecule produced by certain Cyanobacteria to protect against harmful UV radiation, was studied in endolithic cyanobacterial colonies in the halite crust from one of the driest places on Earth – the hyperarid zone of the Atacama Desert. The distribution of the pigment within the evaporitic crust was studied in detail by various independent analytical methods: Raman spectroscopy (including Raman imaging); advanced microscopic observations (fluorescence microscopy, confocal laser scanning microscopy, low-temperature scanning electron microscopy); and spectrophotometric analyses. The differences in scytonemin biosynthesis were mapped within the colonized interior layers, which can be divided into scytonemin-rich and scytonemin-poor zones. A 532 nm laser for excitation proved to be an ideal excitation source with which to observe the relative content of scytonemin within a particular cell aggregate, as well as between different cell aggregates; based on the scytonemin/carotenoid Raman signal intensity ratio of selected corroborative bands for these two compounds. Significantly, scytonemin was found to accumulate within a decayed biomass in the surface portions of the halite crust. These were found to be highly enriched in both the absolute scytonemin content (as documented by UV/VIS spectrophotometry) and its content relative to other pigments associated with the cyanobacterial cells (e.g. carotenoids and chlorophyll).

     

  29. Barahona, S., Dorador, C., Zhang, R. Aguilar, P., Sand, W., Vera, M., Remonsellez, F., (2014), Isolation and characterization of a novel Acidithiobacillus ferrivorans strain from the Chilean Altiplano: Attachment and biofilm formation on pyrite at low temperatura. Research in Microbiology, 165, 782-793.
    [ Show Abstract ]Microorganisms are used to aid the extraction of valuable metals from low-grade sulfide ores in mines worldwide, but relatively little is known about this process in cold environments. This study comprises a preliminary analysis of the bacterial diversity of the polyextremophilic acid River Aroma located in the Chilean Altiplano, and revealed that Betaproteobacteria was the most dominant bacterial group (Gallionella-like and Thiobacillus-like). Taxa characteristic of leaching environments, such Acidithiobacillus and Leptospirillum, were detected at low abundances. Also, bacteria not associated with extremely acidic, metal-rich environments were found. After enrichment in iron- and sulfur-oxidizing media, we isolated and identified a novel psychrotolerant Acidithiobacillus ferrivorans strain ACH. This strain can grow using ferrous iron, sulfur, thiosulfate, tetrathionate and pyrite, as energy sources. Optimal growth was observed in the presence of pyrite, where cultures reached a cell number of 6.5 · 10(7) cells mL(-1). Planktonic cells grown with pyrite showed the presence of extracellular polymeric substances (10 °C and 28 °C), and a high density of cells attached to pyrite grains were observed at 10 °C by electron microscopy. The attachment of cells to pyrite coupons and the presence of capsular polysaccharides were visualized by using epifluorescence microscopy, through nucleic acid and lectin staining with Syto(®)9 and TRITC-Con A, respectively. Interestingly, we observed high cell adhesion including the formation of microcolonies within 21 days of incubation at 4 °C, which was correlated with a clear induction of capsular polysaccharides production. Our data suggests that attachment to pyrite is not temperature-dependent in At. ferrivorans ACH. The results of this study highlight the potential of this novel psychrotolerant strain in oxidation and attachment to minerals under low-temperature conditions.

     

  30. Francisco J. Díaz, Sandra V. Pereda and Alejandro H. Buschmann, (2015), Coexistence in a subtidal habitat in southern Chile: the effects of giant kelp Macrocystis pyrifera overgrowth on the slipper limpet Crepipatella fecunda, Journal of the Marine Biological Association of the United Kingdom, 95, 25-33.
    [ Show Abstract ]In many coastal areas substrate is the limiting resource for benthic organisms. Some sessile species can be used as secondary substrate, reducing competition and increasing coexistence. In southern Chile, annual populations of Macrocystis pyrifera recruit and grow on the shells of Crepipatella fecunda. This study describes ecological interactions between the kelp and the slipper limpet over an annual cycle. The degree of kelp overgrowth was established by collecting sporophytes and through in situ submarine photography over a 10 month period (starting when kelp recruits became visible and ending when sporophytes were no longer present). Changes in the biochemical composition of the limpet tissue were also recorded by chemical analyses, to evaluate the potential effects (positive/neutral/negative) of kelp on C. fecunda nutritional condition. The results indicate that both species coexist, although kelp overgrowth may cause a decrease in carbohydrates in C. fecunda tissues, restricted to the period when the kelp forest reaches its maximum biomass. Individually, the short duration of the maximum overgrowth period and the size reached by C. fecunda females (up to 65 mm shell length) may enable rapid limpet recovery, avoiding competitive exclusion. On a population level, the M. pyrifera annual cycle generates the needed ‘break’ for C. fecunda populations, reducing the effects of kelp overgrowth. Thus, in the view of the neutral effect of kelp overgrowth, together with the positive effect of C. fecunda on M. pyrifera recruitment described somewhere else, this ecological interaction can be categorized as commensalism.

     

  31. Tomás Correa, Alfonso Gutiérrez, Roberto Flores, Alejandro H. Buschmann, Patricio Cornejo and Cristian Bucarey, (2014), Production and economic assessment of giant kelp Macrocystis pyrifera cultivation for abalone feed in the south of Chile, Aquaculture Research, 1-10.
    [ Show Abstract ]Kelp biomass availability for mass abalone cultivation remains a critical issue in Chile. The technical and economic feasibility of a commercial scale kelp farming activity has not been established. This study describes the production and economic results of a pilot scale unit installed in southern Chile. Our results show 25 kg m−1 of production over a 9-month spring-summer period, and 16.2 kg m−1 during the autumn-winter period. These values indicate that a total biomass production of 41.3 kg (wet) m−1 year−1 can be obtained by placing the culture lines at 4 m intervals. High quality animal food-grade plants with a 9% protein content, over 5 m in length were harvested. Sensitivity analysis showed that by cultivating 30–50 ha with a market value of US$ 78 ton−1, a return on investment can be made after the first year.

     

  32. M. Kiwi, P. Moisset De Espanés, I. Rapaport, G. Theyssier, (2014), Strict Majority Bootstrap Percolation in the r-wheel, Information Processing Letters, 114, 277281.
    [ Show Abstract ]In the strict Majority Bootstrap Percolation process each passive vertex v becomes active if at least View the MathML source of its neighbors are active (and thereafter never changes its state). We address the problem of finding graphs for which a small proportion of initial active vertices is likely to eventually make all vertices active. We study the problem on a ring of n vertices augmented with a “central” vertex u. Each vertex in the ring, besides being connected to u, is connected to its r closest neighbors to the left and to the right. We prove that if vertices are initially active with probability p>1/4 then, for large values of r , percolation occurs with probability arbitrarily close to 1 as n→∞. Also, if p<1/4, then the probability of percolation is bounded away from 1. Keywords Bootstrap percolation; Interconnection networks

     

  33. Al-Hafedh, Y.S., A. Alam & A.H. Buschmann, (2014), Bioremediation potential, growth and biomass yield of the green seaweed, Ulva lactuca in an integrated marine aquaculture system at the Red Sea coast of Saudi Arabia at different stocking densities and effluent flow rates. Reviews in Aquaculture, 6, 1-11.
    [ Show Abstract ]Growth, production and biofiltration rates of seaweed, Ulva lactuca were investigated at two stocking densities (3 kg and 6 kg m−2) and two effluent flow rates (5.4 and 10.8 m3 day−1) to optimize an integrated mariculture system at Saudi Red Sea coast. effluents from fish-rearing tank, stocked with 200 kg fish (Oreochromis spilurus), fed to six seaweed tanks via sedimentation tank. Fish growth (weight gain 1.75 g fish day−1), net production (NP, 10.16 kg m−3) and survival (94.24%) were within acceptable limits. Ulva showed significantly higher (F = 62.62, d.f. 3, 35; P < 0.0001) specific growth rates at lower density compared with higher density and under high flow versus low flow (SGR = 5.78% vs. 2.55% at lower flow and 10.60% vs. 6.26% at higher flow). Biomass yield of Ulva at low- and high-stocking densities (111.11 and 83.2 g wet wt m−2 day−1, respectively) at low flow and (267.44 and 244.19 g wet wt m−2 day−1, respectively) at high flow show that high flow rate and lower density favoured growth. Removal rates of total ammonia nitrogen (TAN) (0.26–0.31 g m−2 day−1) and phosphate phosphorus (0.32–0.41 g m−2 day−1) by U. lactuca were not significantly different (F = 1.9, d.f. 3, 59; P = 0.1394 for TAN and F = 0.29, d.f. 3, 59; P = 0.8324 for phosphates) at both the flow rates and stocking densities. Results show that the effluent flow rate has significant impact over the performance of the seaweed than stocking density. Keywords Bioremediation; Oreochromis niloticus; Red Sea; Ulva lactuca

     

  34. Pérez-Fuentes C, Cristina Ravanal M, Eyzaguirre J., (2014), Heterologous expression of a Penicillium purpurogenum pectin lyase in Pichia pastoris and its characterization, Fungal Biology, 118 507-515.
    [ Show Abstract ]Lignocellulose is the major component of plant cell walls and it represents a great source of renewable organic matter. One of lignocellulose constituents is pectin. Pectin is composed of two basic structures: a ‘smooth’ region and a ‘hairy’ region. The ‘smooth’ region (homogalacturonan) is a linear polymer of galacturonic acid residues with α-(1→4) linkages, substituted by methyl and acetyl residues. The ‘hairy’ region is more complex, containing xylogalacturonan and rhamnogalacturonans I and II. Among the enzymes which degrade pectin (pectinases) is pectin lyase (E.C. 4.2.2.10). This enzyme acts on highly esterified homogalacturonan, catalysing the cleavage of α-(1→4) glycosidic bonds between methoxylated residues of galacturonic acid by means of β-elimination, with the
    formation of 4,5-unsaturated products. In this work, the gene and cDNA of a pectin lyase from Penicillium purpurogenum have been sequenced, and the cDNA has been expressed in Pichia pastoris. The gene is 1334 pb long, has three introns and codes for a protein of 376 amino acid residues. The recombinant enzyme was purified to homogeneity and characterized. Pectin lyase has a molecular mass of 45 kDa as determined by SDS-PAGE. It is active on highly esterified pectin, and decreases 40% the viscosity of pectin with a degree of esterification ≥85%. The enzyme showed no activity on polygalacturonic acid and pectin from citrus fruit 8% esterified. The optimum pH and temperature for the recombinant enzyme are 6.0 and 50 °C, respectively, and it is stable up to 50 °C when exposed for 3 h. A purified pectin lyase may be useful in biotechnological applications such as the food industry where the liberation of toxic methanol in pectin degradation should be avoided. Keywords Acidic pectin lyases; Biotechnological applications of pectin lyase; Gene sequencing; Homogalacturonan degradation; Polysaccharide lyase family 1

     

  35. Cristina Ravanal M, Eyzaguirre J., (2015), Heterologous expression and haracterization of α-L-arabinofuranosidase 4 from Penicillium purpurogenum and comparison with the other isoenzymes produced by the fungus, Fungal Biology, 1-7.
    [ Show Abstract ]Penicillium purpurogenum secretes at least four arabinofuranosidases. In this work, the gene of α-L-arabinofuranosidase 4 (ABF4) has been sequenced and expressed in Pichia pastoris. The gene is 1521 pb long, has no introns and codes for a protein of 506 amino acid residues including a signal peptide of 26 residues. Mature protein has a calculated molecular mass of 55.4 kDa, shows 77% identity with α-Larabinofuranosidase 1 from P. purpurogenum and belongs to family 54 of the glycosyl hydrolases. Purified enzyme has a molecular mass near 68 kDa, is active on p-nitrophenyl α-L-arabinofuranoside and p-nitrophenyl-β-D-galactofuranoside, and follows Michaelis-Menten kinetics with KM of 1.58 ± 0.13 mM and 5.3 ± 1.18 mM, respectively. The pH optimum is 4.6 and optimal temperature is 50 °C. The enzyme is active on sugar beet arabinan and wheat flour arabinoxylan but does not act on short arabinooligosaccharides or debranched arabinan. It shows synergistic effect on arabinose liberation from wheat arabinoxylan when combined with endoxylanase from P. purpurogenum. The properties of ABF4 have been compared with those of the other arabinofuranosidases produced by the fungus. P. purpurogenum is the first fungus possessing four biochemically characterized arabinofuranosidases. The availability of four different ABFs may be valuable for biotechnological applications. Keywords Arabinoxylan; GH family 54; Lignocellulose biodegradation; Pichia pastoris; Recombinat enzyme

     

  36. Del-Cid A, Ubilla P, Ravanal MC, Medina E, Vaca I, Levicán G, Eyzaguirre J, Chávez R., (2014), Cold-Active Xylanases Produced by Fungi Associated with Antarctic Marine Sponges, Appl Biochem Biotechnol, 172, 524-523.
    [ Show Abstract ]Despite their potential biotechnological applications, cold-active xylanolytic enzymes have been poorly studied. In this work, 38 fungi isolated from marine sponges collected in King George Island, Antarctica, were screened as new sources of coldactive xylanases. All of them showed xylanase activity at 15 and 23 °C in semiquantitative plate assays. One of these isolates, Cladosporium sp., showed the highest activity and was characterized in detail. Cladosporium sp. showed higher xylanolytic activity when grown on beechwood or birchwood xylan and wheat bran, but wheat straw and oat bran were not so good inducers of this activity. The optimal pH for xylanase activity was 6.0, although pH stability was slightly wider (pH 5-7). On the other hand, Cladosporium sp. showed high xylanase activity at low temperatures and very low thermal stability. Interestingly, thermal stability was even lower after culture media were removed and replaced by buffer, suggesting that low molecular component(s) of the culture media could be important in the stabilization of cold-active xylanase activity. To the best of our knowledge, this study is the first report on extracellular xylanase production by fungi associated with Antarctic marine sponges.