We have applied three machine learning approaches: Random Forest Regression (RFR), Boosted Regression Tree (BRT) and Support Vector Regression (SVR) to the prediction of warfarin maintenance dose in a cohort of African Americans. We have developed a multi-step approach that selects SNPs, builds prediction models with different subsets of selected SNPs along with known associated genetic and environmental variables and tests the discovered models in a cross-validation framework. Preliminary results indicate that our modeling approach gives much higher accuracy than previous models for warfarin dose prediction. A model size of 200 SNPs (in addition to the known genetic and environmental variables) gives the best accuracy. The R(2) between the predicted and actual square root of warfarin dose in this model was on average 66.4% for RFR, 57.8% for SVR and 56.9% for BRT. Thus RFR had the best accuracy, but all three techniques achieved better performance than the current published R(2) of 43% in a sample of mixed ethnicity, and 27% in an African American sample. In summary, machine learning approaches for high-dimensional pharmacogenetic prediction, and for prediction of clinical continuous traits of interest, hold great promise and warrant further research.

No abstract

ObjectivesMany conservative methods have been applied in the treatment of pilonidal sinus disease (PSD). The most commonly used conservative treatment is 80% phenol solution. Our observations demonstrated that 80% phenol solution caused much destruction in the sacrococcygeal region.DesignIn this study low concentrations of phenol were used with the aim of reducing the unwanted side-effects of high-concentration phenol without reducing the therapeutic effects.ParticipantsWe treated 112 patients (18 women, 94 men) with PSD using phenol solution. Patients were divided into two groups: Group A was treated with a 40% solution of phenol solution, and Group B was treated with an 80% solution of phenol solution.SettingAll patients were treated on an outpatient basis. One mL of low (40%) or high (80%) concentration phenol solution was injected into the main sinus orifice. During the check it was observed and noted whether there was skin necrosis, fatty tissue necrosis or abscesses.Main outcome measuresThe mean age was 27.4 years (6–44). The median length of symptoms was seven months (0.5–132). In the 2.8 years (1–6) of mean follow-up period, the disease recurred in 13 (11.6%) patients.ResultsThis treatment procedure was well-tolerated by all the patients except for those who had unwanted results. No patients in group A had skin necrosis, and only one had abscesses. In group B two patients had abscesses, and three had skin necrosis. Fatty tissue necrosis was seen in one patient in Group A and in five patients in Group B. Recurrence rates were four (7.4%) cases in Group A and nine (15.5%) cases in Group B.ConclusionsIt is possible to treat patients in a shorter time with a considerably smaller loss of working time, since the destruction of peripilonidal adipose tissue and skin is less. Therefore, the use of low-concentration phenol solution is an option to be considered in the treatment of PSD.

No abstract

No abstract

OBJECTIVE Recent studies have established that hemispheric diffuse gliomas may be grouped into subsets on the basis of molecular markers; these subsets are loosely correlated with the histopathological diagnosis but are strong predictors of clinical tumor behavior. Based on an analysis of molecular and clinical parameters, the authors hypothesized that mutations of the telomerase promoter (TERTp-mut) mark separate oncogenic programs among isocitrate dehydrogenase 1 and/or 2 (IDH) mutant (IDH-mut) and IDH wild-type (IDH-wt) diffuse gliomas independent of histopathology or WHO grade. METHODS Four molecular subsets of the combined statuses of IDH and TERT-promoter mutations (double mutant, IDH only, TERT only, and double negative) were defined. Differences in age, anatomical location, molecular genetics, and survival rates in a surgical cohort of 299 patients with a total of 356 hemispheric diffuse gliomas (WHO Grade II, III, or IV) were analyzed. RESULTS TERTp-mut were present in 38.8% of IDH-mut and 70.2% of IDH-wt gliomas. The mutational status was stable in each patient at 57 recurrence events over a 2645-month cumulative follow-up period. Among patients with IDH-mut gliomas, those in the double-mutant subset had better survival and a lower incidence of malignant degeneration than those in the IDH-only subset. Of patients in the double-mutant subset, 96.3% were also positive for 1p/19q codeletions. All patients with 1p/19q codeletions had TERTp-mut. In patients with IDH-mut glioma, epidermal growth factor receptor or phosphatase and tensin homolog mutations were not observed, and copy-number variations were uncommon. Among IDH-wt gliomas, the TERT-only subset was associated with significantly higher age, higher Ki-67 labeling index, primary glioblastoma-specific oncogenic changes, and poor survival. The double-negative subset was genetically and biologically heterogeneous. Survival analyses (Kaplan-Meier, multivariate, and regression-tree analyses) confirmed that patients in the 4 molecular subsets had distinct prognoses. CONCLUSIONS Molecular subsets result in different tumor biology and clinical behaviors in hemispheric diffuse gliomas.

Background: Thymic stromal lymphopoietin (TSLP) is expressed by airway epithelial cells and plays a key role in immunological events in asthma. Data on the genetic variants of TSLP and its association with asthma and allergic rhinitis are scarce. We aimed to investigate the effects of the genetic variants of TSLP in children with asthma and allergic rhinitis. Methods: The genetic variants of the TSLP gene were determined by sequencing 25 asthmatic and 25 healthy children. In an association study, a population of 506 asthmatics and 157 healthy controls was screened for the following single-nucleotide polymorphisms (SNPs): rs3806933 and rs2289276 in the promoter region; rs11466741, rs11466742, and rs2289278 in intron 2; rs10073816, rs11466749, and rs11466750 in exon 4, and rs11466754 in 3′-UTR. Results: In Multifactor Dimensionality Reduction analysis, presence of the rs11466749 AA genotype with atopy was significantly associated with a diagnosis of asthma (testing set accuracy: 0.720 and cross validation: 9/10). Two functional SNPs showed a gender-specific association with allergy, i.e. the rs3806933 CC genotype with asthma in boys (p = 0.032, nonsignificant after multiple testing) and the rs2289276 CC genotype with higher eosinophil numbers in asthmatic girls (p = 0.003). The presence of allergic rhinitis in asthmatic children strengthened the association of the rs11466749 GG genotype with asthma (p = 0.001), and rs2289276 was significantly associated with lower FEV₁ levels in asthmatics without allergic rhinitis (p = 0.003). Conclusion: Variants in the gene encoding the TSLP protein may have differential effects on asthma phenotypes depending on gender, atopy, and the presence of allergic rhinitis.

Missing observations are always a challenging problem that we have to deal with in diseases that require follow-up. In hospital records for vesicoureteral reflux (VUR) and recurrent urinary tract infection (rUTI), the number of complete cases is very low on demographic and clinical characteristics, laboratory findings, and imaging data. On the other hand, deep learning (DL) approaches can be used for highly missing observation scenarios with its own missing ratio algorithm. In this study, the effects of multiple imputation techniques MICE and FAMD on the performance of DL in the differential diagnosis were compared. The data of a retrospective cross-sectional study including 611 pediatric patients were evaluated (425 with VUR, 186 with rUTI, 26.65% missing ratio) in this research. CNTK and R 3.6.3 have been used for evaluating different models for 34 features (physical, laboratory, and imaging findings). In the differential diagnosis of VUR and rUTI, the best performance was obtained by deep learning with MICE algorithm with its values, respectively, 64.05% accuracy, 64.59% sensitivity, and 62.62% specificity. FAMD algorithm performed with accuracy=61.52, sensitivity=60.20, and specificity was found out to be 61.00 with 3 principal components on missing imputation phase. DL-based approaches can evaluate datasets without doing preomit/impute missing values from datasets. Once DL method is used together with appropriate missing imputation techniques, it shows higher predictive performance.