Adjusted geometric-mean: a novel performance measure for imbalanced bioinformatics datasets learning

One common and challenging problem faced by many bioinformatics applications, such as promoter recognition, splice site prediction, RNA gene prediction, drug discovery and protein classification, is the imbalance of the available datasets. In most of these applications, the positive data examples are largely outnumbered by the negative data examples, which often leads to the development of sub-optimal prediction models having high negative recognition rate (Specificity = SP) and low positive recognition rate (Sensitivity = SE). When class imbalance learning methods are applied, usually, the SE is increased at the expense of reducing some amount of the SP. In this paper, we point out that in these data-imbalanced bioinformatics applications, the goal of applying class imbalance learning methods would be to increase the SE as high as possible by keeping the reduction of SP as low as possible. We explain that the existing performance measures used in class imbalance learning can still produce sub-optimal models with respect to this classification goal. In order to overcome these problems, we introduce a new performance measure called Adjusted Geometric-mean (AGm). The experimental results obtained on ten real-world imbalanced bioinformatics datasets demonstrates that the AGm metric can achieve a lower rate of reduction of SP than the existing performance metrics, when increasing the SE through class imbalance learning methods. This characteristic of AGm metric makes it more suitable for achieving the proposed classification goal in imbalanced bioinformatics datasets learning.     

Mannose-binding lectin-associated serine protease-1 is a significant contributor to coagulation in a murine model of occlusive thrombosis

Bleeding disorders and thrombotic complications constitute a major cause of death and disability worldwide. Although it is known that the complement and coagulation systems interact, no studies have investigated the specific role or mechanisms of lectin-mediated coagulation in vivo. FeCl(3) treatment resulted in intra-arterial occlusive thrombogenesis within 10 min in wild-type (WT) and C2/factor B-null mice. In contrast, mannose-binding lectin (MBL)-null and MBL-associated serine protease (MASP)-1/-3 knockout (KO) mice had significantly decreased FeCl(3)-induced thrombogenesis. Reconstitution with recombinant human (rh) MBL restored FeCl(3)-induced thrombogenesis in MBL-null mice to levels comparable to WT mice, suggesting a significant role of the MBL/MASP complex for in vivo coagulation. Additionally, whole blood aggregation demonstrated increased MBL/MASP complex-dependent platelet aggregation. In vitro, MBL/MASP complexes were captured on mannan-coated plates, and cleavage of a chromogenic thrombin substrate (S2238) was measured. We observed no significant differences in S2238 cleavage between WT, C2/factor B-null, MBL-A(-/-), or MBL-C(-/-) sera; however, MBL-null or MASP-1/-3 KO mouse sera demonstrated significantly decreased S2238 cleavage. rhMBL alone failed to cleave S2238, but cleavage was restored when rMASP-1 was added to either MASP-1/-3 KO sera or rhMBL. Taken together, these findings indicate that MBL/MASP complexes, and specifically MASP-1, play a key role in thrombus formation in vitro and in vivo.     

Solution structure of chicken liver basic fatty acid binding protein

Chicken liver basic fatty acid binding protein (Lb-FABP) belongs to the basic-type fatty acid binding proteins, a novel group of proteins isolated from liver of different non mammalian species whose structure is not known. The structure of Lb-FABP has been solved by ¹H NMR. The overall fold of Lb-FABP, common to the other proteins of the family, consists of ten antiparallel -strands organised in two nearly ortogonal -sheets with two alpha helices closing the protein cavity where small hydrophobic ligands can be bound. The binding specificity of the protein is not known, however, based on the high sequence and structural similarity with an orthologous protein, ileal lipid binding protein, it is suggested that bile acids may be the putative ligands.     

Characterization of a mammalian Golgi-localized protein complex, COG, that is required for normal Golgi morphology and function

Multiprotein complexes are key determinants of Golgi apparatus structure and its capacity for intracellular transport and glycoprotein modification. Three complexes that have previously been partially characterized include (a) the Golgi transport complex (GTC), identified in an in vitro membrane transport assay, (b) the ldlCp complex, identified in analyses of CHO cell mutants with defects in Golgi-associated glycosylation reactions, and (c) the mammalian Sec34 complex, identified by homology to yeast Sec34p, implicated in vesicular transport. We show that these three complexes are identical and rename them the conserved oligomeric Golgi (COG) complex. The COG complex comprises four previously characterized proteins (Cog1/ldlBp, Cog2/ldlCp, Cog3/Sec34, and Cog5/GTC-90), three homologues of yeast Sec34/35 complex subunits (Cog4, -6, and -8), and a previously unidentified Golgi-associated protein (Cog7). EM of ldlB and ldlC mutants established that COG is required for normal Golgi morphology. "Deep etch" EM of purified COG revealed an approximately 37-nm-long structure comprised of two similarly sized globular domains connected by smaller extensions. Consideration of biochemical and genetic data for mammalian COG and its yeast homologue suggests a model for the subunit distribution within this complex, which plays critical roles in Golgi structure and function.     

Characterization of Bordetella pertussis strains of recent isolation.

During the clinical trial conducted in Italy to evaluate the efficacy of new acellular pertussis vaccines, the most favorable conditions for the recovery and characterization of Bordetella pertussis strains, isolated from children with cough, were adopted. The nasopharyngeal aspirates were collected and sent to the laboratory in refrigerated conditions within 24 hours. Charcoal agar selective and non selective plates were used, and most of the isolates were recovered after 3-4 days of incubation. Confirmation of all suspected colonies included the use of biochemical tests and specific agglutination reaction with B. pertussis antiserum. Serotyping of fimbriae, susceptibility to erythromycin and DNA fingerprinting by Pulsed Field Gel Electrophoresis (PFGE), were performed to characterize B. pertussis isolates and to determine relatedness among different strains. Serotype 1,3 was the most represented in the bacterial population examined. A predominant pulsetype (PTA) characterized most of the isolates accounting for 71.4% of the strains examined. Eight subclones (23.5%) and three unrelated pulsetypes were also found. No resistant strains to erythromycin were detected.     

Difficulties in diagnosing chronic constrictive pericarditis

This presentation calls attention to the many problems involved in the positive, aetiological and differential diagnosis of chronic constrictive pericarditis. We mention the difficulties in aetiological diagnosis in the absence of an episode of acute pericarditis in the past medical history and the clinical findings similar to vascular decompensated cirrhosis or idiopathic restrictive cardiomyopathy. ECG and two-dimensional echocardiography do not have an important role in diagnosis, and in the absence of computed tomography and magnetic resonance imaging, chest radiography, especially a lateral view, could establish the diagnosis. A delay in diagnosis creates difficulties in the surgical treatment, but this treatment improves the patient''s condition in the long term more than the short term.     

Dimeric Smac mimetics/IAP inhibitors as in vivo-active pro-apoptotic agents. Part II: Structural and biological characterization

Novel pro-apoptotic, homodimeric and heterodimeric Smac mimetics/IAPs inhibitors connected through head–head (8), tail–tail (9) or head–tail linkers (10), were biologically and structurally characterized. In vitro characterization (binding to BIR3 and linker-BIR2–BIR3 domains from XIAP and cIAP1, cytotoxicity assays) identified early leads from each dimer family. Computational models and structural studies (crystallography, NMR, gel filtration) partially rationalized the observed properties for each dimer class. Tail–tail dimer 9a was shown to be active in a breast and in an ovary tumor model, highlighting the potential of dimeric Smac mimetics/IAP inhibitors based on the N-AVPI-like 4-substituted 1-aza-2-oxobicyclo[5.3.0]decane scaffold as potential antineoplastic agents.     

Increased coronary arteriolar contraction to serotonin in juvenile pigs with metabolic syndrome

Serum albumin (SA) has been shown to act as a heme scavenger in hemolysis and can protect cell against the toxic effect of heme. However, the mechanism of SA in heme detoxification is not well understood. Interestingly, increasing studies indicate that heme/H₂O₂-dependent reaction is unlikely to be the principal cause of heme toxicity in excessive intravascular hemolysis conditions. Moreover, high levels of NO₂^? and NO₃^? were also found in patients with severe hemolytic diseases, which seem to involve in heme toxic effect as well. Therefore, we proposed that studying the protection mechanism of SA against the heme/H₂O₂/NO₂^?-induced cytotoxicity may be more consistent with free heme-associated disorder pathologies. In this study, we tested the hypotheses that tyrosine residues of bovine serum albumin (BSA) play a prominent role in detoxifying heme in SH-SY5Y cells. Both BSA and tyrosine modified BSA (BSA-T) were used to explore this protective mechanism. Most of cellular injury (oxidative and nitrative damage) induced by heme/H₂O₂/NO₂^? were prevented by pretreatment with an equimolar concentration of BSA or BSA-T, and BSA was found more efficient than BSA-T. Meanwhile, BSA or BSA-T binding to heme is not accompanied by a decrease of heme’s peroxidase activity. Collectively, these data suggest that the protecting effect of BSA against heme-induced damage in the intravascular hemolysis diseases is not accomplished by preventing the primary reactivity of heme with H₂O₂, but by trapping radical through special residues such as tyrosine to render other important protein less damaged.