Prediction of postpartum depression using multilayer perceptrons and pruning.

OBJECTIVE. The main goal of this paper is to obtain a classification model based on feed-forward multilayer perceptrons in order to improve postpartum depression prediction during the 32 weeks after childbirth with a high sensitivity and specificity and to develop a tool to be integrated in a decision support system for clinicians. MATERIALS AND METHODS. Multilayer perceptrons were trained on data from 1397 women who had just given birth, from seven Spanish general hospitals, including clinical, environmental and genetic variables. A prospective cohort study was made just after delivery, at 8 weeks and at 32 weeks after delivery. The models were evaluated with the geometric mean of accuracies using a hold-out strategy. RESULTS. Multilayer perceptrons showed good performance (high sensitivity and specificity) as predictive models for postpartum depression. CONCLUSIONS. The use of these models in a decision support system can be clinically evaluated in future work. The analysis of the models by pruning leads to a qualitative interpretation of the influence of each variable in the interest of clinical protocols.

Dectin-1 and DC-SIGN polymorphisms associated with invasive pulmonary Aspergillosis infection.

The recognition of pathogen-derived structures by C-type lectins and the chemotactic activity mediated by the CCL2/CCR2 axis are critical steps in determining the host immune response to fungi. The present study was designed to investigate whether the presence of single nucleotide polymorphisms (SNPs) within DC-SIGN, Dectin-1, Dectin-2, CCL2 and CCR2 genes influence the risk of developing Invasive Pulmonary Aspergillosis (IPA). Twenty-seven SNPs were selected using a hybrid functional/tagging approach and genotyped in 182 haematological patients, fifty-seven of them diagnosed with proven or probable IPA according to the 2008 EORTC/MSG criteria. Association analysis revealed that carriers of the Dectin-1(rs3901533 T/T) and Dectin-1(rs7309123 G/G) genotypes and DC-SIGN(rs4804800 G), DC-SIGN(rs11465384 T), DC-SIGN(7248637 A) and DC-SIGN(7252229 C) alleles had a significantly increased risk of IPA infection (OR = 5.59 95%CI 1.37-22.77; OR = 4.91 95%CI 1.52-15.89; OR = 2.75 95%CI 1.27-5.95; OR = 2.70 95%CI 1.24-5.90; OR = 2.39 95%CI 1.09-5.22 and OR = 2.05 95%CI 1.00-4.22, respectively). There was also a significantly increased frequency of galactomannan positivity among patients carrying the Dectin-1(rs3901533_T) allele and Dectin-1(rs7309123_G/G) genotype. In addition, healthy individuals with this latter genotype showed a significantly decreased level of Dectin-1 mRNA expression compared to C-allele carriers, suggesting a role of the Dectin-1(rs7309123) polymorphism in determining the levels of Dectin-1 and, consequently, the level of susceptibility to IPA infection. SNP-SNP interaction (epistasis) analysis revealed significant interactions models including SNPs in Dectin-1, Dectin-2, CCL2 and CCR2 genes, with synergistic genetic effects. Although these results need to be further validated in larger cohorts, they suggest that Dectin-1, DC-SIGN, Dectin-2, CCL2 and CCR2 genetic variants influence the risk of IPA infection and might be useful in developing a risk-adapted prophylaxis.