Alterations in the Serum Proteome Following Electroconvulsive Therapy for a Major Depressive Episode: A Longitudinal Multicenter Study.

Background: Electroconvulsive therapy (ECT) is the most effective treatment for severe depression, but the biological changes induced by ECT remain poorly understood. Methods: This study investigated alterations in blood serum proteins in 309 patients receiving ECT for a major depressive episode. We analyzed 201 proteins in samples collected at 3 time points (T): just before the first ECT treatment session (T0), within 30 minutes after the first ECT session (T1), and just before the sixth ECT session (T2). Results: Using statistical models to account for repeated sampling, we identified 152 and 70 significantly (<5% false discovery rate) altered proteins at T1 and T2, respectively. The most pronounced alterations at T1 were transiently increased levels of prolactin, myoglobin, and kallikrein-6. However, most proteins had decreased levels at T1, with the largest effects observed for pro-epidermal growth factor, proto-oncogene tyrosine-protein kinase Src, tumor necrosis factor ligand superfamily member 14, sulfotransferase 1A1, early activation antigen CD69, and CD40 ligand. The change of several acutely altered proteins correlated with electric current and pulse frequency in a dose-response-like manner. Over a 5-session course of ECT, some acutely altered levels were sustained while others increased, e.g., serine protease 8 and chitinase-3-like protein 1. None of the studied protein biomarkers were associated with clinical response to ECT. Conclusions: We report experimental data on alterations in the circulating proteome triggered by ECT in a clinical setting. The findings implicate hormonal signaling, immune response, apoptotic processes, and more. None of the findings were associated with clinical response to ECT.

A loop-counting method for covariate-corrected low-rank biclustering of gene-expression and genome-wide association study data.

A common goal in data-analysis is to sift through a large data-matrix and detect any significant submatrices (i.e., biclusters) that have a low numerical rank. We present a simple algorithm for tackling this biclustering problem. Our algorithm accumulates information about 2-by-2 submatrices (i.e., 'loops') within the data-matrix, and focuses on rows and columns of the data-matrix that participate in an abundance of low-rank loops. We demonstrate, through analysis and numerical-experiments, that this loop-counting method performs well in a variety of scenarios, outperforming simple spectral methods in many situations of interest. Another important feature of our method is that it can easily be modified to account for aspects of experimental design which commonly arise in practice. For example, our algorithm can be modified to correct for controls, categorical- and continuous-covariates, as well as sparsity within the data. We demonstrate these practical features with two examples; the first drawn from gene-expression analysis and the second drawn from a much larger genome-wide-association-study (GWAS).