Selecting high-dimensional mixed graphical models using minimal AIC or BIC forests

Background  

Chow and Liu showed that the maximum likelihood tree for multivariate discrete distributions may be found using a maximum weight spanning tree algorithm, for example Kruskal's algorithm. The efficiency of the algorithm makes it tractable for high-dimensional problems.     

Esters of trehalose from Corynebacterium diphtheriae: A modified purification procedure and studies on the structure of their constituent hydroxylated fatty acids

The purification procedure of 6,6′-diesters of trehalose from Corynebacterium diphtheriae was modified and the isolated substance was analysed by mass spectrometry as its permethylated derivative. The fatty acid moiety released from the glycolipid after alkaline hydrolysis was studied by mass spectral analysis of the O-methylated and O-acetylated methyl ester derivatives. By argentation thin-layer chromatography, three species of O-acetylated methyl esters were recognized, corresponding to saturated, mono-unsaturated and di-unsaturated α-branched-β-hydroxylated fatty acids. The double bond was located by ozonolysis of the O-acetylated methyl ester derivatives, by gas chromatography of the reaction product and mass spectrometry of the effluent from the gas chromatograph. The main components of each species of α-branched-β-hydroxylated fatty acids found in the gly colipid fraction of C. diphtheriae were 2-tetradecyl-3-hydroxyoctadecanoic acid (C₃₂H₆₄O₃, corynomycolic acid), 2-tetradecyl-3-hydroxy-11-octadecenoic acid (C₃₂H₆₂O₃, corynomycolenic acid), 2-tetradec-7′-enyl-3-hydroxy octadecanoic acid (C₃₂H₆₂O₃) and 2-tetradec-7′-enyl-3-hydroxy-11-octadecenoic acid (C₃₂H₆₀O₃, corynomycoldienic acid). The glycolipid fraction from C. diphtheriae is obviously a complex mixture of 6,6′-diesters of trehalose.     

Production, crystallization, and preliminary X-ray analysis of rabbit skeletal muscle troponin complex consisting of troponin C and fragment (1-47) of troponin I.

Troponin is a ternary protein complex consisting of subunits TnC. TnI, and TnT, and plays a key role in calcium regulation of the skeletal and cardiac muscle contraction. In the present study, a partial complex (CI47) was prepared from Escherichia coli-expressed rabbit skeletal muscle TnC and fragment 1-47 of TnI, which is obtained by chemical cleavage of an E. coli-expressed mutant of rabbit skeletal muscle TnI. Within the ternary troponin complex, CI47 is thought to form a core that is resistant to proteolytic digestion, and the interaction within CI47 likely maintains the integrity of the troponin complex. Complex CI47 was crystallized in the presence of sodium citrate. The addition of trehalose improved the diffraction pattern of the crystals substantially. The crystal lattice belongs to the space group P3(1)(2)21, with unit cell dimensions a = b = 48.2 A, c = 162 A. The asymmetric unit presumably contains one CI47 complex. Soaking with p-chloromercuribenzenesulfonate (PCMBS) resulted in loss of isomorphism, but enhanced the quality of the crystals. The crystals diffracted up to 2.3 A resolution, with completeness of 91% and R(merge) = 6.4%. The crystals of PCMBS-derivative should be suitable for X-ray studies using the multiple-wavelength anomalous diffraction technique. This is the first step for elucidating the structure of the full troponin complex.     

Techniques for the detection of pathogenic Cryptococcus species in wood decay substrata and the evaluation of viability in stored samples

In this study, we evaluated several techniques for the detection of the yeast form of Cryptococcus in decaying wood and measured the viability of these fungi in environmental samples stored in the laboratory. Samples were collected from a tree known to be positive for Cryptococcus and were each inoculated on 10 Niger seed agar (NSA) plates. The conventional technique (CT) yielded a greater number of positive samples and indicated a higher fungal density [in colony forming units per gram of wood (CFU.g^-1) ] compared to the humid swab technique (ST). However, the difference in positive and false negative results between the CT-ST was not significant. The threshold of detection for the CT was 0.05.10³ CFU.g^-1, while the threshold for the ST was greater than 0.1.10³ CFU^-1. No colonies were recovered using the dry swab technique. We also determined the viability of Cryptococcus in wood samples stored for 45 days at 25ºC using the CT and ST and found that samples not only continued to yield a positive response, but also exhibited an increase in CFU.g^-1, suggesting that Cryptococcus is able to grow in stored environmental samples. The ST.1, in which samples collected with swabs were immediately plated on NSA medium, was more efficient and less laborious than either the CT or ST and required approximately 10 min to perform; however, additional studies are needed to validate this technique.     

Mucosal layers and related nerve fibres in non-chagasic and chagasic human colon—a quantitative immunohistochemical study

Chagasic megacolon is accompanied by extensive myenteric and, simultaneously, moderate submucosal neuron loss. Here, we examined changes of the innervation pattern of the lamina propria (LP) and muscularis mucosae (MM). Two alternating sets of cryosections were taken from seven non-chagasic colonic and seven chagasic megacolonic specimens (the latter included both the dilated megacolonic and the non-dilated transitional oral and anal zones) and were immunohistochemically triple-stained for smooth-muscle actin (SMA), synaptophysin (SYN) and glial acid protein S100 and, alternatively, for SMA, vasoactive intestinal peptide (VIP) and somatostatin (SOM). Subsequent image analysis and statistical evaluation of nervous tissue profile areas revealed that, in LP, the most extreme differences (i.e. increase in thickness or decrease in nerve, glia and muscle tissue profile area, respectively) compared with control values occurred in the dilated megacolonic zone itself. In contrast, the most extreme differences in the MM were in the anal-to-megacolonic zone (except the profile area of muscle tissue, which was lowest in the megacolonic zone). This parallels our previous results in the external muscle coat. A partial and selective survival of VIP-immunoreactive in contrast to SOM-immunoreactive nerve fibres was observed in both mucosal layers investigated. Thus, VIPergic nerve elements might be crucial for the maintenance of the mucosal barrier. The differential changes of neural tissue parameters in LP and MM might reflect a multifactorial rather than a pure neurogenic development of megacolon in chronic Chagas’ disease.