Model selection and Bayesian methods in statistical genetics: summary of group 11 contributions to Genetic Analysis Workshop 15

The research presented in group 11 of the Genetic Analysis Workshop 15 (GAW15) falls into two major themes: Model selection approaches for gene mapping (both Bayesian and Frequentist); and other Bayesian methods. These methods either allow relaxation of some of the common assumptions, such as mode of inheritance, for studying complicated genetic systems, or allow incorporation of additional information into the model. Over half of the groups applied model selection methods on all three data sets, using models in which genetic markers were used as predictors for linkage, phenotype expression, or transmission to an affected offspring. Most groups employed variations of Stochastic Search Variable Selection as the model selection method of choice. A brief review of this class of methods is given in this summary paper, followed by highlights of other methods and overall summaries of each contribution to the GAW15 presentation group 11. These group contributions exhibit the value of framing genetic problems in terms of model selection, and highlight the impact of variable selection for gene mapping.     

Motivation and value influences in the relative balance of goal-directed and habitual behaviours in obsessive-compulsive disorder

Our decisions are based on parallel and competing systems of goal-directed and habitual learning, systems which can be impaired in pathological behaviours. Here we focus on the influence of motivation and compare reward and loss outcomes in subjects with obsessive-compulsive disorder (OCD) on model-based goal-directed and model-free habitual behaviours using the two-step task. We further investigate the relationship with acquisition learning using a one-step probabilistic learning task. Forty-eight OCD subjects and 96 healthy volunteers were tested on a reward and 30 OCD subjects and 53 healthy volunteers on the loss version of the two-step task. Thirty-six OCD subjects and 72 healthy volunteers were also tested on a one-step reversal task. OCD subjects compared with healthy volunteers were less goal oriented (model-based) and more habitual (model-free) to reward outcomes with a shift towards greater model-based and lower habitual choices to loss outcomes. OCD subjects also had enhanced acquisition learning to loss outcomes on the one-step task, which correlated with goal-directed learning in the two-step task. OCD subjects had greater stay behaviours or perseveration in the one-step task irrespective of outcome. Compulsion severity was correlated with habitual learning in the reward condition. Obsession severity was correlated with greater switching after loss outcomes. In healthy volunteers, we further show that greater reward magnitudes are associated with a shift towards greater goal-directed learning further emphasizing the role of outcome salience. Our results highlight an important influence of motivation on learning processes in OCD and suggest that distinct clinical strategies based on valence may be warranted.     

Effects of age-dependent membrane transport changes on the homeostasis of senescent human red blood cells

Little is known about age-related changes in red blood cell (RBC) membrane transport and homeostasis. We investigated first whether the known large variation in plasma membrane Ca(2+) (PMCA) pump activity was correlated with RBC age. Glycated hemoglobin, Hb A1c, was used as a reliable age marker for normal RBCs. We found an inverse correlation between PMCA strength and Hb A1c content, indicating that PMCA activity declines monotonically with RBC age. The previously described subpopulation of high-Na(+), low-density RBCs had the highest Hb A1c levels, suggesting it represents a late homeostatic condition of senescent RBCs. Thus, the normal densification process of RBCs with age must undergo late reversal, requiring a membrane permeability increase with net NaCl gain exceeding KCl loss. Activation of a nonselective cation channel, Pcat, was considered the key link in this density reversal. Investigation of Pcat properties showed that its most powerful activator was increased intracellular Ca(2+). Pcat was comparably selective to Na(+), K(+), choline, and N-methyl-D-glucamine, indicating a fairly large, poorly selective cation permeability pathway. Based on these observations, a working hypothesis is proposed to explain the mechanism of progressive RBC densification with age and of the late reversal to a low-density condition with altered ionic gradients.     

Outcome of children with nodular lymphocyte predominant Hodgkin lymphoma - a Children's Cancer and Leukaemia Group report

This report describes the clinical outcomes and follow-up records of 42 children with nodular lymphocyte predominant Hodgkin lymphoma (LPHL) treated on United Kingdom Children's Cancer Study Group (UKCCSG) HD1 (1982-1992) and HD2 protocols (1992-2000). The clinical records of 42 children with LPHL treated between 1982 and 2000 were reviewed retrospectively. All 42 had histology reviewed centrally and confirmed as LPHL by an expert panel. In both trials, only patients with stage IA disease had the option of being treated with either involved field radiation alone or combination chemotherapy consisting of chlorambucil, vinblastine, procarbazine and prednisolone (ChlVPP). Patients with all other stages were treated with ChlVPP chemotherapy. Thirty-five patients (83%) presented with early stage disease (Stages I & II). All 42 patients achieved a complete remission (CR). Six children relapsed after primary therapy. The 5- and 10-year relapse-free survival rates were 87% and 82% respectively. Forty-one are currently alive in CR. In conclusion, children with low-stage LPHL treated between 1982 and 2000 according to the UK strategy for classical Hodgkin lymphoma (HL) had an excellent prognosis. There have been no second malignancies or transformations to B-cell non-Hodgkin lymphoma.     

Endothelin-converting enzyme 1 promotes re-sensitization of neurokinin 1 receptor-dependent neurogenic inflammation

Background and purpose:

The metalloendopeptidase endothelin-converting enzyme 1 (ECE-1) is prominently expressed in the endothelium where it converts big endothelin to endothelin-1, a vasoconstrictor peptide. Although ECE-1 is found in endosomes in endothelial cells, the role of endosomal ECE-1 is unclear. ECE-1 degrades the pro-inflammatory neuropeptide substance P (SP) in endosomes to promote recycling and re-sensitization of its neurokinin 1 (NK₁) receptor. We investigated whether ECE-1 regulates NK₁ receptor re-sensitization and the pro-inflammatory effects of SP in the endothelium.

Experimental approach:

We examined ECE-1 expression, SP trafficking and NK₁ receptor re-sensitization in human microvascular endothelial cells (HMEC-1), and investigated re-sensitization of SP-induced plasma extravasation in rats.

Key results:

HMEC-1 expressed all four ECE-1 isoforms (a-d), and fluorescent SP trafficked to early endosomes containing ECE-1b/d. The ECE-1 inhibitor SM-19712 prevented re-sensitization of SP-induced Ca²⁺ signals in HMEC-1 cells. Immunoreactive ECE-1 and NK₁ receptors co-localized in microvascular endothelial cells in the rat. SP-induced extravasation of Evans blue in the urinary bladder, skin and ears of the rat desensitized when the interval between two SP injections was 10 min, and re-sensitized after 480 min. SM-19712 inhibited this re-sensitization.

Conclusions and implications:

By degrading endocytosed SP, ECE-1 promotes the recycling and re-sensitization of NK₁ receptors in endothelial cells, and thereby induces re-sensitization of the pro-inflammatory effects of SP. Thus, ECE-1 inhibitors may ameliorate the pro-inflammatory actions of SP.     

A potential role for Dkk-1 in the pathogenesis of osteosarcoma predicts novel diagnostic and treatment strategies

Canonical Wnt signalling is an osteoinductive signal that promotes bone repair through acceleration of osteogenic differentiation by progenitors. Dkk-1 is a secreted inhibitor of canonical Wnt signalling and thus inhibits osteogenesis. To examine a potential osteoinhibitory role of Dkk-1 in osteosarcoma (OS), we measured serum Dkk-1 in paediatric patients with OS (median age, 13.4 years) and found it to be significantly elevated. We also found that Dkk-1 was maximally expressed by the OS cells at the tumour periphery and in vitro, Dkk-1 and RANKL are coexpressed by rapidly proliferating OS cells. Both Dkk-1 and conditioned media from OS cells reduce osteogenesis by human mesenchymal cells and by immunodepletion of Dkk-1, or by adding a GSK3beta inhibitor, the effects of Dkk-1 were attenuated. In mice, we found that the expression of Dkk-1 from implanted tumours was similar to the human tumour biopsies in that human Dkk-1 was present in the serum of recipient animals. These data demonstrate that systemic levels of Dkk-1 are elevated in OS. Furthermore, the expression of Dkk-1 by the OS cells at the periphery of the tumour probably contributes to its expansion by inhibiting repair of the surrounding bone. These data demonstrate that Dkk-1 may serve as a prognostic or diagnostic marker for evaluation of OS and furthermore, immunodepletion of Dkk-1 or administration of GSK3beta inhibitors could represent an adjunct therapy for this disease.     

A dual role for prediction error in associative learning

Confronted with a rich sensory environment, the brain must learn statistical regularities across sensory domains to construct causal models of the world. Here, we used functional magnetic resonance imaging and dynamic causal modeling (DCM) to furnish neurophysiological evidence that statistical associations are learnt, even when task-irrelevant. Subjects performed an audio-visual target-detection task while being exposed to distractor stimuli. Unknown to them, auditory distractors predicted the presence or absence of subsequent visual distractors. We modeled incidental learning of these associations using a Rescorla-Wagner (RW) model. Activity in primary visual cortex and putamen reflected learning-dependent surprise: these areas responded progressively more to unpredicted, and progressively less to predicted visual stimuli. Critically, this prediction-error response was observed even when the absence of a visual stimulus was surprising. We investigated the underlying mechanism by embedding the RW model into a DCM to show that auditory to visual connectivity changed significantly over time as a function of prediction error. Thus, consistent with predictive coding models of perception, associative learning is mediated by prediction-error dependent changes in connectivity. These results posit a dual role for prediction-error in encoding surprise and driving associative plasticity.     

Immune heparin-induced thrombocytopenia resulting from preceding exposure to heparin catheter flushes

Immune heparin-induced thrombocytopenia (HIT) is a life-threatening adverse effect of heparin. It can result from any type of heparin exposure and by any route of administration; however only a few cases are reported after exposures to small quantities of heparin from catheter flushes. The major clinical problem associated with HIT is thrombosis. Early detection and institution of alternative, non-heparin anticoagulation are important. We report a patient with HIT associated with use of therapeutic-dose unfractionated heparin in whom immune sensitization to heparin was triggered by two 500-unit exposure to UFH associated with intravascular catheter flushing for antineoplastic chemotherapy in a patient with colon adenocarcinoma.