Breaking the hierarchy - a new cluster selection mechanism for hierarchical clustering methods

Background  

Hierarchical clustering methods like Ward's method have been used since decades to understand biological and chemical data sets. In order to get a partition of the data set, it is necessary to choose an optimal level of the hierarchy by a so-called level selection algorithm. In 2005, a new kind of hierarchical clustering method was introduced by Palla et al. that differs in two ways from Ward's method: it can be used on data on which no full similarity matrix is defined and it can produce overlapping clusters, i.e., allow for multiple membership of items in clusters. These features are optimal for biological and chemical data sets but until now no level selection algorithm has been published for this method.     

The Semiglades: The Collision of Restoration,Social Values,and the Ecosystem Concept

Defining success targets in restoration and how social values affect them are two commonly discussed issues in restoration today. We believe that how success is commonly defined—with vague terms such as “healthy ecosystem” or cited as a return to a previous, historic state—needs to be reevaluated. With the increasing number of novel ecosystems, there is an increasing conflict between the ecosystem concept, social values, and restoration. This arises from the fact that ecosystems are defined by the values of the scientists describing them, necessarily constraining the ecosystem to a generally static concept. It is not directly the concept, but how it is perceived through our filter of social values that represses the creativity and innovation needed in restoration today. Within restoration, we feel that the ecosystem concept does a disservice by ignoring the increasing number of novel systems, and that hinders real progress in a time when hesitation can be costly. To best illustrate this, we offer the example of restoration of the Florida Everglades and how it has become a novel system in pattern and process. We suggest renaming the Everglades “The Semiglades” in hopes of opening a dialog to expose social/ecosystem biases and include novel landscapes in management and planning.     

The early lethality of autosomal monosomy in the mouse

Using male mice doubly heterozygous for pairs of Robertsonian translocation chromosomes that have one arm in common, mouse embryos monosomic for 11 of the 19 autosomes have been generated. All of these monosomies result in death prior to or during the implantation period, with only rare survivors being detected 6 days after fertilization. For some of the monosomies the onset of lethality can be detected during the third or fourth day of development, but others do not begin to die until sometime after the late blastocyst stage on day 4. Retardation of development, as revealed by decreased cell numbers, is often detectable prior to or after the onset of the lethal period. The period during which death occurs may spread over several days and does not coincide with any of the developmental landmarks of the pre- or peri-implantation period. Genetic factors that may affect the rate of cellular proliferation or other aspects of embryonic development appear to play an important role in determining exactly when individual monosomies result in death. The universal early lethality of the autosomal monosomies leads to the conclusion that a large number of loci scattered over all of the autosomes are involved in processes that are so concentration dependent that a 50% reduction is sufficient to produce very serious consequences.     

Analysis of simian virus 40 wild-type and mutant virions by agarose gel electrophoresis.

Intact wild-type simian virus 40 particles can be separated and resolved from a temperature-sensitive mutant and from a number of other viruses by agarose gel electrophoresis. The relative mobilities of the viruses appear to be a function of both virion size and surface composition. The virions of a temperature-sensitive strain of simian virus 40, tsB204, have significantly greater mobility than those of wild-type simian virus 40, when electrophoresis was conducted toward the cathode at pH 5.0. When electrophoresis was performed toward the anode at pH 7.0, TSB204 viruses have a slightly slower mobility as compared with that of the wild type. The data indicated that the virions of tsB204 have a greater positive charge at their surface than those of wild-type particles. No differences were detected in the finger print patterns of the tryptic peptides of VP1 and VP3 of these two virus strains. Although it was not possible to identify the structural polypeptide directly affected by the tsB204 mutation, we have shown that this mutation affects charge distribution on the surface of the virion.     

Alteration in cell permeability as a mechanism of action of certain quinone pesticides

The permeability of the Chlorella pyrenoidosa membrane was studied by following the efflux of ¹⁴C-intracellular material from cells which had been allowed to incorporate ¹⁴CO₂ photosynthetically. It was observed that the efflux increased upon treatment with low concentrations (3-30 μM) of 2, 3-dichloro-1, 4-naphthoquinone (dichlone), 2-amino-3-chloro-1, 4-naphthoquinone (06K-quinone), and 2, 3, 5, 6-tetrachloro-1, 4-benzoquinone (chloranil). Dichlone caused a greater loss of intracellular material than chloranil or 06K-quinone. The rate of loss as well as the total loss of ¹⁴C increased with an increase in the concentration of the quinones. In the dichlone-treated cells, the leakage was observed within 1 minute of the addition of the chemical and the effect on cell permeability was irreversible. Cells exposed to dichlone in the light or under anaerobic conditions released significantly greater amounts of ¹⁴C-material than cells treated in the dark or under aerobic conditions. The aqueous ethanol-soluble fraction of the cell was found to be the source of the released material. The proportion of the ethanol-soluble ¹⁴C that leaked out of the cell varied with the time of ¹⁴C-assimilation prior to treatment with dichlone. In the dichlone-treated cells, practically all the ¹⁴C-sucrose, alanine, glutamine, serine, and glycine leaked out, whereas glutamic, aspartic, succinic, and fumaric acids were lost only partially. Essentially no ¹⁴C-lipids were lost from the cells during dichlone treatment.     

Herpes simplex virus type 2 glycoprotein gF and type 1 glycoprotein gC have related antigenic determinants.

The 104-S monoclonal antibody immunoprecipitated from herpes simplex virus type 2 (HSV-2)-infected cell extracts the 75,000-molecular-weight glycoprotein gF and its 65,000-molecular-weight precursor (pgF). The precursor pgF was sensitive to endoglycosidase H digestion, indicating the presence of high mannose-type oligosaccharides, whereas the stable gF product was sensitive to neuraminidase digestion, indicating the presence of sialic acid residues. The 104-S antibody also weakly precipitated the 130,000-molecular-weight herpes simplex virus type 1 (HSV-1) glycoprotein gC from both infected cell extracts and purified preparations obtained through the use of monoclonal antibody-containing immunoadsorbent columns. Immunofluorescence tests demonstrated that the 104-S antibody reacted with antigen present in cells infected with HSV-2 strain 333 and HSV-1 strain 14012 but not with antigen present in cells infected with HSV-1 strain MP, a strain deficient in HSV-1 gC production. These findings indicate that HSV-1 gC and HSV-2 gF have antigenic determinants that are related.     

The two components of spectrin, filamin, and the heavy chain of smooth muscle myosin show no detectable homologies to one another by two-dimensional mapping of iodinated tryptic peptides.

The possible structural relationships among four high molecular weight mechanochemical proteins has been examined using two-dimensional mapping of the tryptic peptide fragments prepared from ¹²⁵I-labeled proteins (Elder et al., J. Biol. Chem. $\text{252}$:6510–6515 (1977)). Erythrocyte spectrin bands 1 and 2 protein, the heavy chain of smooth muscle (uterine) myosin, a filamin from human and rabbit were studied. The maps of the four proteins within each species differed considerably from each other, with no apparent homologies evident among them, whereas maps of the same individual protein between the two species showed a high degree of homology.