A nondegenerate penalized likelihood estimator for variance parameters in multilevel models

Group-level variance estimates of zero often arise when fitting multilevel or hierarchical linear models, especially when the number of groups is small. For situations where zero variances are implausible a priori, we propose a maximum penalized likelihood approach to avoid such boundary estimates. This approach is equivalent to estimating variance parameters by their posterior mode, given a weakly informative prior distribution. By choosing the penalty from the log-gamma family with shape parameter greater than 1, we ensure that the estimated variance will be positive. We suggest a default log-gamma(2,λ) penalty with λ → 0, which ensures that the maximum penalized likelihood estimate is approximately one standard error from zero when the maximum likelihood estimate is zero, thus remaining consistent with the data while being nondegenerate. We also show that the maximum penalized likelihood estimator with this default penalty is a good approximation to the posterior median obtained under a noninformative prior.Our default method provides better estimates of model parameters and standard errors than the maximum likelihood or the restricted maximum likelihood estimators. The log-gamma family can also be used to convey substantive prior information. In either case-pure penalization or prior information-our recommended procedure gives nondegenerate estimates and in the limit coincides with maximum likelihood as the number of groups increases.

Pam3Cys4 internalization by naïve CD4+T cells accelerates t-bet promoter epigenetic modification and restrains TGF−β1 expressing tumor growth (P2153)

To promote Th1 responses TLR2 engagement is conducted on ex vivo-activated T cells but this hinders Ag-specific clonal expansion in vivo. We asked if naive CD4+T cells could internalize the TLR2 agonist Pam3Cys4 to promote Th1 responses in vivo. Methods: WT, TLR2-/- or OT2 CD69-CD25-CD44-CD62L+CD4+T cells were incubated with rhodamine-labeled or unlabeled Pam3Cys4 for 3 hrs and extensively washed to remove all soluble agonist. CD4+T(iPam) activation ex vivo was conducted under Th1 conditions +/- TGF-β1 and bisulfite conversion t-bet promoter sequence analysis was performed. 4x10(5) OT2 or OT2(iPam) cells were adoptively transferred into B6 mice with established EG-7 OVA tumors. Results: Intracellular FACS revealed WT or TLR2-/- CD4+T cells readily internalize Pam3Cys4. As compared to untreated CD4+T cells CD4+T(iPam) displayed > than 50% t-bet promoter CpG hypomethylation at -134 and -75 at 36 hrs post-activation leading to sharply higher t-bet and IFN-γ expression in a TLR2 dependent manner. OT2 (iPam) halted EG-7 OVA growth with high numbers of OVA-specific CD4+ and CD8+ T cells in the DLN and tumor. In contrast, untreated OT2 induced little Th1 accumulation and accelerated tumor growth when compared to mice that did not receive T cells but this rapid growth pattern could be reversed with TGF-β1 Abs. Also, unlike with untreated OT2 cells TGF-β1 could not block naïve OT2(iPam) differentiation into Th1 cells. These data suggest a novel approach to tumor immunotherapy.

Loss of homeostatic granulopoiesis in IKKβ-deficient mice (P4472)

BACKGROUND IL-1β-mediated neutrophilia has been observed in mice with an inducible deletion of IKKβ under non-homeostatic conditions. We investigated the effects of a constitutive IKKβ deletion from the hematopoietic compartment. METHODS Mice carrying loxP-flanked IKKβ alleles (IKKβfl/fl) were crossed with Vav1-iCre mice to target IKKβ expression in hematopoietic cells (IKKβΔ/Δ). Bone marrow (BM), blood, spleen and peripheral tissues were analyzed by histopathology and FACS. Methylcellulose colony-forming cell (CFU) assays were performed using Lin-c-kit+ myeloid progenitors (MP) and splenocytes. MP proliferation was measured by EdU incorporation. RESULTS IKKβΔ/fl mice are phenotypically indistinguishable from IKKβfl/fl (WT) mice. In contrast, IKKβΔ/Δ mice are born in non-Mendelian ratios, develop patchy alopecia, osteopetrosis, splenomegaly and stunted growth. IKKβΔ/Δ mice also show massive neutrophilia in BM, spleen, blood and peripheral tissues as well as a sharp increase in BM and spleen granulocyte CFUs relative to IKKβΔ/fl and WT controls. Unlike in IKKβΔ/fl or WT granulocyte macrophage progenitors (GMP), IKKβΔ/Δ GMPs proliferate more in response to G-CSF than to GM-CSF or IL-3 stimulation. Moreover, G-CSF stimulated IKKβΔ/Δ MP produce more neutrophils relative to IKKβΔ/fl and WT MP. CONCLUSIONS Our data indicate that IKKβ controls homeostatic granulopoiesis through regulating sensitivity to G-CSF signals in MP.

Rates and correlates of HIV and STI infection among homeless women

We studied the prevalence of biologically confirmed HIV, Chlamydia, and gonorrhea in a randomly selected sample of sheltered homeless women in New York City, and explored their association with demographic, homeless history, and clinical risk factors. 329 women were randomly selected from 28 family and single adult shelters. The estimated prevalence of HIV in the study sample is 0.6 % (±0.3 %); for Chlamydia it is 6.7 % (±2.2 %); for gonorrhea it is 0.9 % (±0.04 %). A history of childhood sexual abuse, arrest history, current psychotic symptoms, and substance use disorder placed women at greater risk of infection. We consider contextual factors that may yield underestimates of HIV prevalence in our sample and discuss how a more comprehensive prevalence estimate might be constructed. Findings underscore the importance of offering HIV/STI testing, counseling, and HIV risk prevention interventions to homeless women and suggest that interventions should be tailored to the needs of specific subgroups of homeless women.

Philosophy and the practice of Bayesian statistics

A substantial school in the philosophy of science identifies Bayesian inference with inductive inference and even rationality as such, and seems to be strengthened by the rise and practical success of Bayesian statistics. We argue that the most successful forms of Bayesian statistics do not actually support that particular philosophy but rather accord much better with sophisticated forms of hypothetico-deductivism. We examine the actual role played by prior distributions in Bayesian models, and the crucial aspects of model checking and model revision, which fall outside the scope of Bayesian confirmation theory. We draw on the literature on the consistency of Bayesian updating and also on our experience of applied work in social science. Clarity about these matters should benefit not just philosophy of science, but also statistical practice. At best, the inductivist view has encouraged researchers to fit and compare models without checking them; at worst, theorists have actively discouraged practitioners from performing model checking because it does not fit into their framework.

Philosophy and the practice of Bayesian statistics

A substantial school in the philosophy of science identifies Bayesian inference with inductive inference and even rationality as such, and seems to be strengthened by the rise and practical success of Bayesian statistics. We argue that the most successful forms of Bayesian statistics do not actually support that particular philosophy but rather accord much better with sophisticated forms of hypothetico-deductivism. We examine the actual role played by prior distributions in Bayesian models, and the crucial aspects of model checking and model revision, which fall outside the scope of Bayesian confirmation theory. We draw on the literature on the consistency of Bayesian updating and also on our experience of applied work in social science. Clarity about these matters should benefit not just philosophy of science, but also statistical practice. At best, the inductivist view has encouraged researchers to fit and compare models without checking them; at worst, theorists have actively discouraged practitioners from performing model checking because it does not fit into their framework.

IKKβ in recipient myeloid cells promotes lung allograft acceptance and prevents the accumulation of intragraft IL-17+ CD4+ T cells (169.33)

Background Mechanisms linking acute solid organ injury to rejection are poorly understood. We previously reported maintenance of IKKβ activity in recipient myeloid cells protects lung grafts from acute injury. We now ask if recipient myeloid cell IKKβ activity regulates lung allograft survival. Methods Balb/c lungs were transplanted into B6 recipients carrying A) Myeloid-specific IKKβ deletion (IKKβΔMye) or B) floxed IKKβ alleles (IKKβfl/fl). Post-surgery, recipients were immunosuppressed (IS) with MR1/anti-CD40L and CTLA4-Ig. In a second cohort, IS-IKKβΔMye and IS-IKKβfl/fl recipients were also treated with control Ig, IL-23p19 or IL-17-specific Abs. Lung allografts were analyzed on POD7 for histopathology and accumulation of IFNγ+ and IL-17+ T cells by FACS analysis. Results Acute rejection occurred in IS-IKKβΔMye, but not IS-IKKβfl/fl recipients. Moreover, IS-IKKβΔMye recipients developed marked intragraft accumulation of IFNγ+CD4+, IFNγ+CD8+ and IL-17+CD4+ T cells. IS-IKKβΔMye recipients treated with IL-23p19 Abs accepted their lungs and had fewer intragraft IL-17+CD4+ T cells with little change in IFNγ+CD4+ and IFNγ+CD8+ T cell abundance. By contrast, IL-17 Ab treatment depleted intragraft IL17+CD4+ T cells, but led to expansion of intragraft Th1 cells and did not prevent rejection. Conclusions IKKβ-regulated signals in recipient myeloid cells couple acute lung graft injury to alloimmunity by preventing Th17-mediated rejection in a IL-23p19-dependent manner.