COmic: convolutional kernel networks for interpretable end-to-end learning on (multi-)omics data.

MOTIVATION: The size of available omics datasets is steadily increasing with technological advancement in recent years. While this increase in sample size can be used to improve the performance of relevant prediction tasks in healthcare, models that are optimized for large datasets usually operate as black boxes. In high-stakes scenarios, like healthcare, using a black-box model poses safety and security issues. Without an explanation about molecular factors and phenotypes that affected the prediction, healthcare providers are left with no choice but to blindly trust the models. We propose a new type of artificial neural network, named Convolutional Omics Kernel Network (COmic). By combining convolutional kernel networks with pathway-induced kernels, our method enables robust and interpretable end-to-end learning on omics datasets ranging in size from a few hundred to several hundreds of thousands of samples. Furthermore, COmic can be easily adapted to utilize multiomics data. RESULTS: We evaluated the performance capabilities of COmic on six different breast cancer cohorts. Additionally, we trained COmic models on multiomics data using the METABRIC cohort. Our models performed either better or similar to competitors on both tasks. We show how the use of pathway-induced Laplacian kernels opens the black-box nature of neural networks and results in intrinsically interpretable models that eliminate the need for post hoc explanation models. AVAILABILITY AND IMPLEMENTATION: Datasets, labels, and pathway-induced graph Laplacians used for the single-omics tasks can be downloaded at https://ibm.ent.box.com/s/ac2ilhyn7xjj27r0xiwtom4crccuobst/folder/48027287036. While datasets and graph Laplacians for the METABRIC cohort can be downloaded from the above mentioned repository, the labels have to be downloaded from cBioPortal at https://www.cbioportal.org/study/clinicalData?id=brca\_metabric. COmic source code as well as all scripts necessary to reproduce the experiments and analysis are publicly available at https://github.com/jditz/comics.

PlasmoFAB: a benchmark to foster machine learning for Plasmodium falciparum protein antigen candidate prediction.

MOTIVATION: Machine learning methods can be used to support scientific discovery in healthcare-related research fields. However, these methods can only be reliably used if they can be trained on high-quality and curated datasets. Currently, no such dataset for the exploration of Plasmodium falciparum protein antigen candidates exists. The parasite P.falciparum causes the infectious disease malaria. Thus, identifying potential antigens is of utmost importance for the development of antimalarial drugs and vaccines. Since exploring antigen candidates experimentally is an expensive and time-consuming process, applying machine learning methods to support this process has the potential to accelerate the development of drugs and vaccines, which are needed for fighting and controlling malaria. RESULTS: We developed PlasmoFAB, a curated benchmark that can be used to train machine learning methods for the exploration of P.falciparum protein antigen candidates. We combined an extensive literature search with domain expertise to create high-quality labels for P.falciparum specific proteins that distinguish between antigen candidates and intracellular proteins. Additionally, we used our benchmark to compare different well-known prediction models and available protein localization prediction services on the task of identifying protein antigen candidates. We show that available general-purpose services are unable to provide sufficient performance on identifying protein antigen candidates and are outperformed by our models that were trained on this tailored data. AVAILABILITY AND IMPLEMENTATION: PlasmoFAB is publicly available on Zenodo with DOI 10.5281/zenodo.7433087. Furthermore, all scripts that were used in the creation of PlasmoFAB and the training and evaluation of machine learning models are open source and publicly available on GitHub here: https://github.com/msmdev/PlasmoFAB.

Inherently interpretable position-aware convolutional motif kernel networks for biological sequencing data.

Artificial neural networks show promising performance in detecting correlations within data that are associated with specific outcomes. However, the black-box nature of such models can hinder the knowledge advancement in research fields by obscuring the decision process and preventing scientist to fully conceptualize predicted outcomes. Furthermore, domain experts like healthcare providers need explainable predictions to assess whether a predicted outcome can be trusted in high stakes scenarios and to help them integrating a model into their own routine. Therefore, interpretable models play a crucial role for the incorporation of machine learning into high stakes scenarios like healthcare. In this paper we introduce Convolutional Motif Kernel Networks, a neural network architecture that involves learning a feature representation within a subspace of the reproducing kernel Hilbert space of the position-aware motif kernel function. The resulting model enables to directly interpret and evaluate prediction outcomes by providing a biologically and medically meaningful explanation without the need for additional post-hoc analysis. We show that our model is able to robustly learn on small datasets and reaches state-of-the-art performance on relevant healthcare prediction tasks. Our proposed method can be utilized on DNA and protein sequences. Furthermore, we show that the proposed method learns biologically meaningful concepts directly from data using an end-to-end learning scheme.

Perturbation-evoked potentials can be classified from single-trial EEG.

OBJECTIVE: Loss of balance control can have serious consequences on interaction between humans and machines as well as the general well-being of humans. Perceived balance perturbations are always accompanied by a specific cortical activation, the so-called perturbation-evoked potential (PEP). In this study, we investigate the possibility to classify PEPs from ongoing EEG. APPROACH: Fifteen healthy subjects were exposed to seated whole-body perturbations. Each participant performed 120 trials; they were rapidly tilted to the right and left, 60 times respectively. MAIN RESULTS: We achieved classification accuracies of more than 85% between PEPs and rest EEG using a window-based classification approach. Different window lengths and electrode layouts were compared. We were able to achieve excellent classification performance (87.6 ± 8.0% accuracy) by using a short window length of 200 ms and a minimal electrode layout consisting of only the Cz electrode. The peak classification accuracy coincides in time with the strongest component of PEPs, called N1. SIGNIFICANCE: We showed that PEPs can be discriminated against ongoing EEG with high accuracy. These findings can contribute to the development of a system that can detect balance perturbations online.

Properties of Organic Liquids when Simulated with Long-Range Lennard-Jones Interactions.

In order to increase the accuracy of classical computer simulations, existing methodologies may need to be adapted. Hitherto, most force fields employ a truncated potential function to model van der Waals interactions, sometimes augmented with an analytical correction. Although such corrections are accurate for homogeneous systems with a long cutoff, they should not be used in inherently inhomogeneous systems such as biomolecular and interface systems. For such cases, a variant of the particle mesh Ewald algorithm (Lennard-Jones PME) was already proposed 20 years ago (Essmann et al. J. Chem. Phys. 1995, 103, 8577-8593), but it was implemented only recently (Wennberg et al. J. Chem. Theory Comput. 2013, 9, 3527-3537) in a major simulation code (GROMACS). The availability of this method allows surface tensions of liquids as well as bulk properties to be established, such as density and enthalpy of vaporization, without approximations due to truncation. Here, we report on simulations of ≈150 liquids (taken from a force field benchmark: Caleman et al. J. Chem. Theory Comput. 2012, 8, 61-74) using three different force fields and compare simulations with and without explicit long-range van der Waals interactions. We find that the density and enthalpy of vaporization increase for most liquids using the generalized Amber force field (GAFF, Wang et al. J. Comput. Chem. 2004, 25, 1157-1174) and the Charmm generalized force field (CGenFF, Vanommeslaeghe et al. J. Comput. Chem. 2010, 31, 671-690) but less so for OPLS/AA (Jorgensen and Tirado-Rives, Proc. Natl. Acad. Sci. U.S.A. 2005, 102, 6665-6670), which was parametrized with an analytical correction to the van der Waals potential. The surface tension increases by ≈10(-2) N/m for all force fields. These results suggest that van der Waals attractions in force fields are too strong, in particular for the GAFF and CGenFF. In addition to the simulation results, we introduce a new version of a web server, http://virtualchemistry.org, aimed at facilitating sharing and reuse of input files for molecular simulations.