A lower bound on the size of k-neighborhood in generalized cubes

The determination of the minimum size of a k-neighborhood (i.e., a neighborhood of a set of k nodes) in a given graph is essential in the analysis of diagnosability and fault tolerance of multicomputer systems. The generalized cubes include the hypercube and most hypercube variants as special cases. In this paper, we present a lower bound on the size of a k-neighborhood in n-dimensional generalized cubes, where 2n + 1 <= k <= 3n - 2. This lower bound is tight in that it is met by the n-dimensional hypercube. Our result is an extension of two previously known results. (c) 2005 Elsevier Inc. All rights reserved.

Dynamic load balancing in parallel KD-tree k-means

One among the most influential and popular data mining methods is the k-Means algorithm for cluster analysis. Techniques for improving the efficiency of k-Means have been largely explored in two main directions. The amount of computation can be significantly reduced by adopting geometrical constraints and an efficient data structure, notably a multidimensional binary search tree (KD-Tree). These techniques allow to reduce the number of distance computations the algorithm performs at each iteration. A second direction is parallel processing, where data and computation loads are distributed over many processing nodes. However, little work has been done to provide a parallel formulation of the efficient sequential techniques based on KD-Trees. Such approaches are expected to have an irregular distribution of computation load and can suffer from load imbalance. This issue has so far limited the adoption of these efficient k-Means variants in parallel computing environments. In this work, we provide a parallel formulation of the KD-Tree based k-Means algorithm for distributed memory systems and address its load balancing issue. Three solutions have been developed and tested. Two approaches are based on a static partitioning of the data set and a third solution incorporates a dynamic load balancing policy.

Carbon monoxide mediates the anti-apoptotic effects of heme oxygenase-1 in medulloblastoma DAOY cells via K+ channel inhibition

Tumor cell survival and proliferation is attributable in part to suppression of apoptotic pathways, yet the mechanisms by which cancer cells resist apoptosis are not fully understood. Many cancer cells constitutively express heme oxygenase-1 (HO-1), which catabolizes heme to generate biliverdin, Fe(2+), and carbon monoxide (CO). These breakdown products may play a role in the ability of cancer cells to suppress apoptotic signals. K(+) channels also play a crucial role in apoptosis, permitting K(+) efflux which is required to initiate caspase activation. Here, we demonstrate that HO-1 is constitutively expressed in human medulloblastoma tissue, and can be induced in the medulloblastoma cell line DAOY either chemically or by hypoxia. Induction of HO-1 markedly increases the resistance of DAOY cells to oxidant-induced apoptosis. This effect was mimicked by exogenous application of the heme degradation product CO. Furthermore we demonstrate the presence of the pro-apoptotic K(+) channel, Kv2.1, in both human medulloblastoma tissue and DAOY cells. CO inhibited the voltage-gated K(+) currents in DAOY cells, and largely reversed the oxidant-induced increase in K(+) channel activity. p38 MAPK inhibition prevented the oxidant-induced increase of K(+) channel activity in DAOY cells, and enhanced their resistance to apoptosis. Our findings suggest that CO-mediated inhibition of K(+) channels represents an important mechanism by which HO-1 can increase the resistance to apoptosis of medulloblastoma cells, and support the idea that HO-1 inhibition may enhance the effectiveness of current chemo- and radiotherapies.

Scalability of efficient parallel K-Means

Clustering is defined as the grouping of similar items in a set, and is an important process within the field of data mining. As the amount of data for various applications continues to increase, in terms of its size and dimensionality, it is necessary to have efficient clustering methods. A popular clustering algorithm is K-Means, which adopts a greedy approach to produce a set of K-clusters with associated centres of mass, and uses a squared error distortion measure to determine convergence. Methods for improving the efficiency of K-Means have been largely explored in two main directions. The amount of computation can be significantly reduced by adopting a more efficient data structure, notably a multi-dimensional binary search tree (KD-Tree) to store either centroids or data points. A second direction is parallel processing, where data and computation loads are distributed over many processing nodes. However, little work has been done to provide a parallel formulation of the efficient sequential techniques based on KD-Trees. Such approaches are expected to have an irregular distribution of computation load and can suffer from load imbalance. This issue has so far limited the adoption of these efficient K-Means techniques in parallel computational environments. In this work, we provide a parallel formulation for the KD-Tree based K-Means algorithm and address its load balancing issues.

Space partitioning for scalable K-means

K-Means is a popular clustering algorithm which adopts an iterative refinement procedure to determine data partitions and to compute their associated centres of mass, called centroids. The straightforward implementation of the algorithm is often referred to as `brute force' since it computes a proximity measure from each data point to each centroid at every iteration of the K-Means process. Efficient implementations of the K-Means algorithm have been predominantly based on multi-dimensional binary search trees (KD-Trees). A combination of an efficient data structure and geometrical constraints allow to reduce the number of distance computations required at each iteration. In this work we present a general space partitioning approach for improving the efficiency and the scalability of the K-Means algorithm. We propose to adopt approximate hierarchical clustering methods to generate binary space partitioning trees in contrast to KD-Trees. In the experimental analysis, we have tested the performance of the proposed Binary Space Partitioning K-Means (BSP-KM) when a divisive clustering algorithm is used. We have carried out extensive experimental tests to compare the proposed approach to the one based on KD-Trees (KD-KM) in a wide range of the parameters space. BSP-KM is more scalable than KDKM, while keeping the deterministic nature of the `brute force' algorithm. In particular, the proposed space partitioning approach has shown to overcome the well-known limitation of KD-Trees in high-dimensional spaces and can also be adopted to improve the efficiency of other algorithms in which KD-Trees have been used.

Factors affecting potassium leaching in different soils

A set of lysimeter based experiments was carried out during 2000/01 to evaluate the impact of soil type and grassland management on potassium (K) leaching. The effects of (1) four soil textures (sand, loam, loam over chalk and clay), (2) grazing and cutting (with farmyard manure application), and (3) K applied as inorganic fertilizer, dairy slurry or a mixture of both sources were tested. Total K losses in the clay soil were more than twice those in the sand soil (13 and 6 kg K ha(-1), respectively) because of the development of preferential flow in the clay soil. They were also greater in the cut treatment than in the grazed treatment (82 and 51 kg K ha(-1), respectively; P less than or equal to0.01), associated with a 63% increase of K concentration in the leachates from the former (6.7 +/- 0.28 and 4.1 +/- 0.22 mg K L-1 for cut and grazed, respectively; P less than or equal to0.01) because of the K input from the farmyard manure. The source of fertilizer did not affect total K losses or the average K concentration in the leachates (P >0.05), but it changed the pattern of these over time.

Uneven distribution of nutrients in the root zone affects the incidence of blossom end rot and concentration of calcium and potassium in fruits of tomato

Tomato plants ( Lycopersicon esculentum Mill. var. DRK) were grown hydroponically to determine the effect of an uneven distribution of nutrients in the root zone on blossomend rot (BER) and Ca and K concentrations in the fruits. The plants were grown in rockwool with their root system divided into two portions. Each portion was irrigated with nutrient solutions with either the same or the different electrical conductivity (EC) in the range 0 to 6 dS m(-1). Solutions with high EC supplied to both sides of the root system significantly increased the incidence of BER. However, when only water or a solution of low EC was supplied to one portion, BER was reduced by 80%. Fruit yields were significantly higher ( P < 0.01) for plants that received solutions of the uneven EC treatments (6/0 or 4.5/0 EC treatment). Plants supplied with solutions of uneven EC generally had higher leaf and fruit concentrations of Ca but lower concentrations of K than those supplied with solutions of high EC. There was no difference in Ca concentration at the distal end of young fruits of the uneven EC treatment but it was reduced in the high EC treatments. The concentration of K in the mature fruits of the uneven EC treatments was lower than that of the high EC treatments and higher or similar that of the 3/3 or 2.5/2.5 EC treatments ( controls). A clear relationship was found between the incidence of BER and the exudation rate. High rate of xylem exudation was observed in the uneven EC treatments. Reduction of BER in the uneven EC treatments is most likely to be the effect of high exudation rate on Ca status in the young fruits. It was concluded that high EC of solution had positive effects on Ca concentration and incidence of BER provided that nutrient solution with low EC or water is supplied to the one portion of the root system.

In vitro degradability of mature and immature leaves of tropical forage legumes differing in condensed tannin and non-starch polysaccharide content and composition

A study was designed to examine the relationships between protein, condensed tannin and cell wall carbohydrate content and composition and the nutritional quality of seven tropical legumes (Desmodium ovalifolium, Flemingia macrophylla, Leucaena leucocephala, L pallida, L macrophylla, Calliandra calothyrsus and Clitotia fairchildiana). Among the legume species studied, D ovalifolium showed the lowest concentration of nitrogen, while L leucocephala showed the highest. Fibre (NDF) content was lowest in C calothyrsus, L Leucocephala and L pallida and highest in L macrophylla, which had no measurable condensed tannins. The highest tannin concentration was found in C calothyrsus. Total non-structural polysaccharides (NSP) varied among legumes species (lowest in C calothyrsus and highest in D ovalifolium), and glucose and uronic acids were the most abundant carbohydrate constituents in all legumes. Total NSP losses were lowest in F macrophylla and highest in L leucocephala and L pallida. Gas accumulation and acetate and propionate levels were 50% less with F macrophylla and D ovalifolium as compared with L leucocephala. The highest levels of branched-chain fatty acids were observed with non-tanniniferous legumes, and negative concentrations were observed with some of the legumes with high tannin content (D ovalifolium and F macrophylla). Linear regression analysis showed that the presence of condensed tannins was more related to a reduction of the initial rate of gas production (0-48 h) than to the final amount of gas produced or the extent (144h) of dry matter degradation, which could be due to differences in tannin chemistry. Consequently, more attention should be given in the future to elucidating the impact of tannin structure on the nutritional quality of tropical forage legumes. (C) 2003 Society of Chemical Industry.

Effect of plasma insulin and branched-chain amino acids on skeletal muscle protein synthesis in fasted lambs

The increase in fractional rate of protein synthesis (K-s) in the skeletal muscle of growing rats during the transition from fasted to fed state has been explained by the synergistic action of a rise in plasma insulin and branched-chain amino acids (BCAA). Since growing lambs Also exhibit an increase in K-s with level of feed intake, the objective of the present study was to determine if this synergistic relationship between insulin and BCAA also occurs in ruminant animals. Six 30 kg fasted (72 h) lambs (8 months of age) received each of four treatments, which were based on continuous infusion into the jugular vein for 6 h of: (1) saline (155 mmol NaCl/l); (2) a mixture of BCAA (0.778 mumol leucine, 0.640 mumol isoleucine and 0.693 mumol valine/min.kg); (3) 18.7 mumol glucose/min.kg (to induce endogenous insulin secretion): (4) co-infusion of BCAA and glucose. Within each period all animals received the same isotope of phenylalanine, (Phe) as follows: (1) L-[1-C-13]Phe; (2) L-phenyl-[ring H-2(5)]-alanine; (3) L-[N-15]Phe; (4) L-[ring 2,6-H-3]Phe. Blood was sampled serially during infusions to measure plasma concentrations of insulin, glucose and amino acids, and plasma free Phe isotopic activity; biopsies were taken 6 h after the beginning of infusions to determine K-s in in. longissimus dorsi and vastus muscle. Compared with control (saline-infused) lambs, K-s was increased by an average of 40% at the end of glucose infusion, but this effect was not statistically significant in either of the muscles sampled. BCAA infusion, alone or in combination with glucose, also had no significant effect on K-s compared with control sheep. K-s was approximately 60% greater for vastus muscle than for m. longissimus dorsi (P<0.01), regardless of treatment. It is concluded that there are signals other than insulin and BCAA that are responsible for the feed-induced increase in K-s in muscle of growing ruminant animals.

Uneven distribution of nutrients in the root zone affects the incidence of blossom end rot and concentration of calcium and potassium in fruits of tomato

Tomato plants ( Lycopersicon esculentum Mill. var. DRK) were grown hydroponically to determine the effect of an uneven distribution of nutrients in the root zone on blossomend rot (BER) and Ca and K concentrations in the fruits. The plants were grown in rockwool with their root system divided into two portions. Each portion was irrigated with nutrient solutions with either the same or the different electrical conductivity (EC) in the range 0 to 6 dS m(-1). Solutions with high EC supplied to both sides of the root system significantly increased the incidence of BER. However, when only water or a solution of low EC was supplied to one portion, BER was reduced by 80%. Fruit yields were significantly higher ( P < 0.01) for plants that received solutions of the uneven EC treatments (6/0 or 4.5/0 EC treatment). Plants supplied with solutions of uneven EC generally had higher leaf and fruit concentrations of Ca but lower concentrations of K than those supplied with solutions of high EC. There was no difference in Ca concentration at the distal end of young fruits of the uneven EC treatment but it was reduced in the high EC treatments. The concentration of K in the mature fruits of the uneven EC treatments was lower than that of the high EC treatments and higher or similar that of the 3/3 or 2.5/2.5 EC treatments ( controls). A clear relationship was found between the incidence of BER and the exudation rate. High rate of xylem exudation was observed in the uneven EC treatments. Reduction of BER in the uneven EC treatments is most likely to be the effect of high exudation rate on Ca status in the young fruits. It was concluded that high EC of solution had positive effects on Ca concentration and incidence of BER provided that nutrient solution with low EC or water is supplied to the one portion of the root system.

In vivo transformations of dihydroartemisinic acid in Artemisia annua plants

[15-(CH3)-C-13-H-2]-dihydroartemisinic acid (2a) and [15-(CH3)-H-2]-dihydroartemisinic acid (2b) have been fed via the root to intact Artemisia annua plants and their transformations studied in vivo by one-dimensional H-2 NMR spectroscopy and two-dimensional, C-13-H-2 correlation NMR spectroscopy (C-13-(2) H COSY). Labelled dihydroartemisinic acid was transformed into 16 12-carboxy-amorphane and cadinane sesquiterpenes within a few days in the aerial parts of A. annua, although transformations in the root were much slower and more limited. Fifteen of these 16 metabolites have been reported previously as natural products from A. annua. Evidence is presented that the first step in the transformation of dihydroartemisinic acid in vivo is the formation of allylic hydroperoxides by the reaction of molecular oxygen with the Delta(4,5)-double bond in this compound. The origin of all 16 secondary metabolites might then be explained by the known further reactions of such hydroperoxides. The qualitative pattern for the transformations of dihydroartemisinic acid in vivo was essentially unaltered when a comparison was made between plants, which had been kept alive and plants which were allowed to die after feeding of the labelled precursor. This, coupled with the observation that the pattern of transformations of 2 in vivo demonstrated very close parallels with the spontaneous autoxidation chemistry for 2, which we have recently demonstrated in vitro, has lead us to conclude that the main 'metabolic route' for dihydroartemisinic acid in A. annua involves its spontaneous autoxidation and the subsequent spontaneous reactions of allylic hydroperoxides which are derived from 2. There may be no need to invoke the participation of enzymes in any of the later biogenetic steps leading to all 16 of the labelled 11,13-dihydro-amorphane sesquiterpenes which are found in A. annua as natural products. (C) 2003 Elsevier Ltd. All rights reserved.

In vivo transformations of artemisinic acid in Artemisia annua plants

Artemisinic acid labeled with both C-13 and H-2 at the 15-position has been fed to intact plants of Artemisia annua via the cut stem, and its in vivo transformations studied by 1D- and 2D-NMR spectroscopy. Seven labeled metabolites have been isolated, all of which are known as natural products from this species. The transformations of artemisinic acid-as observed both for a group of plants, which was kept alive by hydroponic administration of water and for a group, which was allowed to die by desiccation-closely paralleled those, which have been recently described for its 11,13-dihydro analog, dihydroartemisinic acid. It seems likely therefore that similar mechanisms, involving spontaneous autoxidation of the Delta(4,5) double bond in both artemisinic acid and dihydroartemisinic acid and subsequent rearrangements of the resultant allylic hydroperoxides, may be involved in the biological transformations, which are undergone by both compounds. All of the sesquiterpene metabolites, which were obtained from in vivo transformations of artemisinic acid retained their unsaturation at the 11,13-position, and there was no evidence for conversion into any 11,13-dihydro metabolite, including artemisinin, the antimalarial drug, which is produced by A. annua. This observation led to the proposal of a unified biosynthetic scheme, which accounts for the biogenesis of many of the amorphane and cadinane sesquiterpenes that have been isolated as natural products from A. annua. In this scheme, there is a bifurcation in the biosynthetic pathway starting from amorpha-4,11-diene leading to either artemisinic acid or dihydroartemisinic acid; these two committed precursors are then, respectively, the parents for the two large families of highly oxygenated 11,13-dehydro and 11,13-dihydro sesquiterpene metabolites, which are known from this species. (C) 2007 Elsevier Ltd. All rights reserved.

In vivo transformations of dihydro-epi-deoxyarteannuin B in Artemisia annua plants

[15-(CH3)-C-13-H-2]-dihydro-epi-deoxyarteannuin B (4a) has been fed to intact Artemisia annua plants via the root and three labeled metabolites (17a-19a) have been identified by 1D- and 2D-NMR spectroscopies. The in vivo transformations of 4a in A. annua are proposed to involve enzymatically-mediated processes in addition to possible spontaneous autoxidation. In the hypothetical spontaneous autoxidation pathway, the tri-substituted double bond in 4a appears to have undergone 'ene-type' reaction with oxygen to form an allylic hydroperoxide, which subsequently rearranges to the allylic hydroxyl group in the metabolite 3 alpha-hydroxy-dihydro-epi-deoxyarteannuin B (17a). In the enzymatically-mediated pathways, compound 17a has then been converted to its acetyl derivative, 3 alpha-acetoxy-dihydro-epi-deoxyarteannuin B (18a), while oxidation of 4a at the 'unactivated' 9-position has yielded 9 beta-hydroxy-dihydro-epi-deoxyarteannuin B (19a). Although all of the natural products artemisinin ( 1), arteannuin K ( 7), arteannuin L ( 8), and arteannuin M ( 9) have been suggested previously as hypothetical metabolites from dihydro-epi-deoxyarteannuin B in A. annua, none were isolated in labeled form in this study. It is argued that the nature of the transformations undergone by compound 4a are more consistent with a degradative metabolism, designed to eliminate this compound from the plant, rather than with a role as a late precursor in the biosynthesis of artemisinin or other natural products from A. annua. (C) 2007 Elsevier Ltd. All rights reserved.