SummaryMeasuring the similarity of graphs is a fundamental step in the analysis of graph-structured data, which is omnipresent in computational biology. Graph kernels have been proposed as a powerful and efficient approach to this problem of graph comparison. Here we provide graphkernels, the first R and Python graph kernel libraries including baseline kernels such as label histogram based kernels, classic graph kernels such as random walk based kernels, and the state-of-the-art Weisfeiler-Lehman graph kernel. The core of all graph kernels is implemented in C ++ for efficiency. Using the kernel matrices computed by the package, we can easily perform tasks such as classification, regression and clustering on graph-structured samples.Availability and implementationThe R and Python packages including source code are available at https://CRAN.R-project.org/package=graphkernels and https://pypi.python.org/pypi/graphkernels.Supplementary information Supplementary data are available online at Bioinformatics.

Major depressive disorder (MDD) is often accompanied by severe impairments in working memory (WM). Neuroimaging studies investigating the mechanisms underlying these impairments have produced conflicting results. It remains unclear whether MDD patients show hyper- or hypoactivity in WM-related brain regions and how potential aberrations in WM processing may contribute to the characteristic dysregulation of cognition-emotion interactions implicated in the maintenance of the disorder. In order to shed light on these questions and to overcome limitations of previous studies, we applied a multivoxel pattern classification approach to investigate brain activity in large samples of MDD patients (N = 57) and matched healthy controls (N = 61) during a WM task that incorporated positive, negative, and neutral stimuli. Results showed that patients can be distinguished from healthy controls with good classification accuracy based on functional activation patterns. ROI analyses based on the classification weight maps showed that during WM, patients had higher activity in the left DLPFC and the dorsal ACC. Furthermore, regions of the default-mode network (DMN) were less deactivated in patients. As no performance differences were observed, we conclude that patients required more effort, indexed by more activity in WM-related regions, to successfully perform the task. This increased effort might be related to difficulties in suppressing task-irrelevant information reflected by reduced deactivation of regions within the DMN. Effects were most pronounced for negative and neutral stimuli, thus pointing toward important implications of aberrations in WM processes in cognition-emotion interactions in MDD.