Comprehensible Predictive Modeling Using Regularized Logistic Regression and Comorbidity Based Features

Different studies have demonstrated the importance of comorbidities to better understand the origin and evolution of medical complications. This study focuses on improvement of the predictive model interpretability based on simple logical features representing comorbidities. We use group lasso based feature interaction discovery followed by a post-processing step, where simple logic terms are added. In the final step, we reduce the feature set by applying lasso logistic regression to obtain a compact set of non-zero coefficients that represent a more comprehensible predictive model. The effectiveness of the proposed approach was demonstrated on a pediatric hospital discharge dataset that was used to build a readmission risk estimation model. The evaluation of the proposed method demonstrates a reduction of the initial set of features in a regression model by 72%, with a slight improvement in the Area Under the ROC Curve metric from 0.763 (95% CI: 0.755–0.771) to 0.769 (95% CI: 0.761–0.777). Additionally, our results show improvement in comprehensibility of the final predictive model using simple comorbidity based terms for logistic regression.     

Evoked potentials in artificial neural nets

A neural net model of discrete populations of formal neurons have been constructed to study evoked potentials based on our previous simulation studies (Anninos, 1972). Some interesting results came up from the examination of our findings regarding the latencies and the period of the cyclic activity of the evoked potentials. In fact, the different successive latencies for the five identical stimuli and the different periods for each of the cyclic activities, all are consequences of inhibitory and excitatory influences from a large neuronal population. Furthermore, such behavior of the system is not only related to the unknown neuronal population but was also substantially altered by what occurred in other systems at the time of stimulus, or prior to it.Computation assistance was provided by the Health Sciences Computing Facility, UCLA, sponsored by NIH Special Research Resources Grant RR-3. Research was sponsored by NSF Grant GB 30498 and NIH Grant NS-8498.     

Conditioned Medium from Early-Outgrowth Bone Marrow Cells Is Retinal Protective in Experimental Model of Diabetes

Bone marrow-derived cells were demonstrated to improve organ function, but the lack of cell retention within injured organs suggests that the protective effects are due to factors released by the cells. Herein, we tested cell therapy using early outgrowth cells (EOCs) or their conditioned media (CM) to protect the retina of diabetic animal models (type 1 and type 2) and assessed the mechanisms by in vitro study. Control and diabetic (db/db) mice (8 weeks of age) were randomized to receive a unique intravenous injection of 5×10⁵EOCs or 0.25 ml thrice weekly tail-vein injections of 10x concentrated CM and Wystar Kyoto rats rendered diabetic were randomized to receive 0.50 ml thrice weekly tail-vein injections of 10x concentrated CM. Four weeks later, the animals were euthanized and the eyes were enucleated. Rat retinal Müller cells (rMCs) were exposed for 24 h to high glucose (HG), combined or not with EOC-conditioned medium (EOC-CM) from db/m EOC cultures. Diabetic animals showed increase in diabetic retinopathy (DR) and oxidative damage markers; the treatment with EOCs or CM infusions significantly reduced this damage and re-established the retinal function. In rMCs exposed to diabetic milieu conditions (HG), the presence of EOC-CM reduced reactive oxygen species production by modulating the NADPH-oxidase 4 system, thus upregulating SIRT1 activity and deacetylating Lys-310-p65-NFκB, decreasing GFAP and VEGF expressions. The antioxidant capacity of EOC-CM led to the prevention of carbonylation and nitrosylation posttranslational modifications on the SIRT1 molecule, preserving its activity. The pivotal role of SIRT1 on the mode of action of EOCs or their CM was also demonstrated on diabetic retina. These findings suggest that EOCs are effective as a form of systemic delivery for preventing the early molecular markers of DR and its conditioned medium is equally protective revealing a novel possibility for cell-free therapy for the treatment of DR.     

Transcriptional Cofactor TBLR1 Controls Lipid Mobilization in White Adipose Tissue

1. Download : Download high-res image (175KB)
2. Download : Download full-size image

Highlights? TBLR1 controls cAMP-dependent lipolysis in adipocytes ? Adipocyte-specific deletion of TBLR1 in mice impairs fasting-induced lipolysis ? Lack of TBLR1 in adipocytes aggravates diet-induced obesity and metabolic dysfunction ? TBLR1 mRNA levels in WAT are elevated under lipolytic conditions in mice and humans     

Biggest challenges in bioinformatics

The third Heidelberg Unseminars in Bioinformatics (HUB) was held on 18th October 2012, at Heidelberg University, Germany. HUB brought together around 40 bioinformaticians from academia and industry to discuss the ‘Biggest Challenges in Bioinformatics’ in a ‘World Café’ style event.     

Inhibition of intestinal ischemia/repurfusion induced apoptosis and necrosis via down-regulation of the NF-kB,c-Jun and caspace-3 expression by epigallocatechin-3-gallate administration

Intestinal ischemia/reperfusion (I/R) produces reactive oxygen species (ROS) activating signal transduction and apoptosis. The aim of this study was to evaluate the effect of (?)-epigallocatechin-3-gallate (EGCG) administration in inhibition of apoptosis by attenuating the expression of NF-kB, c-Jun and caspace-3 in intestinal I/R. Thirty male wistar rats were used. Group A sham operation, B I/R, C I/R-EGCG 50 mg/kg ip. Intestinal ischemia was induced for 60 min by clamping the superior mesenteric artery. Malondialdehyde (MDA), myeloperoxidase (MPO), light histology, Fragment End Labelling of DNA (TUNEL), immunocytochemistry for NF-kB, c-Jun and caspace-3 analysis in intestinal specimens were performed 120 min after reperfusion. Apoptosis as indicated by TUNEL and Caspace-3, NF-kB and c-Jun was widely expressed in I/R group but only slightly expressed in EGCG treated groups. MDA and MPO showed a marked increase in the I/R group and a significant decrease in the EGCG treated group. Light histology showed preservation of architecture in the EGCG treated group. In conclusion, EGCG pre-treatment is likely to inhibit intestinal I/R-induced apoptosis by down-regulating the expression of NF-kB, c-Jun and caspase-3.     

Synthesis and evaluation of potent and selective MGL inhibitors as a glaucoma treatment

Monoacylglycerol lipase (MGL) inhibition provides a potential treatment approach to glaucoma through the regulation of ocular 2-arachidonoylglycerol (2-AG) levels and the activation of CB1 receptors. Herein, we report the discovery of new series of carbamates as highly potent and selective MGL inhibitors. The new inhibitors showed potent nanomolar inhibitory activity against recombinant human and purified rat MGL, were selective (>1000-fold) against serine hydrolases FAAH and ABHD6 and lacked any affinity for the cannabinoid receptors CB1 and CB2. Protein-based ¹H NMR experiments indicated that inhibitor 2 rapidly formed a covalent adduct with MGL with a residence time of about 6?h. This interconversion process “intrinsic reversibility” was exploited by modifications of the ligand’s size (length and bulkiness) to generate analogs with “tunable’ adduct residence time (τ). Inhibitor 2 was evaluated in a normotensive murine model for assessing intraocular pressure (IOP), which could lead to glaucoma, a major cause of blindness. Inhibitor 2 was found to decrease ocular pressure by ～4.5?mmHg in a sustained manner for at least 12?h after a single ocular application, underscoring the potential for topically-administered MGL inhibitors as a novel therapeutic target for the treatment of glaucoma.     

Enhancement of endocannabinoid signaling by fatty acid amide hydrolase inhibition: A neuroprotective therapeutic modality

AimsThis review posits that fatty acid amide hydrolase (FAAH) inhibition has therapeutic potential against neuropathological states including traumatic brain injury; Alzheimer's, Huntington's, and Parkinson's diseases; and stroke.Main methodsThis proposition is supported by data from numerous in vitro and in vivo experiments establishing metabolic and pharmacological contexts for the neuroprotective role of the endogenous cannabinoid (“endocannabinoid”) system and selective FAAH inhibitors.Key findingsThe systems biology of endocannabinoid signaling involves two main cannabinoid receptors, the principal endocannabinoid lipid mediators N-arachidonoylethanolamine (“anandamide”) (AEA) and 2-arachidonoyl glycerol (2-AG), related metabolites, and the proteins involved in endocannabinoid biosynthesis, biotransformation, and transit. The endocannabinoid system is capable of activating distinct signaling pathways on-demand in response to pathogenic events or stimuli, thereby enhancing cell survival and promoting tissue repair. Accumulating data suggest that endocannabinoid system modulation at discrete targets is a promising pharmacotherapeutic strategy for treating various medical conditions. In particular, neuronal injury activates cannabinoid signaling in the central nervous system as an intrinsic neuroprotective response. Indirect potentiation of this salutary response through pharmacological inhibition of FAAH, an endocannabinoid-deactivating enzyme, and consequent activation of signaling pathways downstream from cannabinoid receptors have been shown to promote neuronal maintenance and function.SignificanceThis therapeutic modality has the potential to offer site- and event-specific neuroprotection under conditions where endocannabinoids are being produced as part of a physiological protective mechanism. In contrast, direct application of cannabinoid receptor agonists to the central nervous system may activate CB receptors indiscriminately and invite unwanted psychotrophic effects.