GNN-SubNet: disease subnetwork detection with explainable Graph Neural Networks

The tremendous success of graphical neural networks (GNNs) has already had a major impact on systems biology research. For example, GNNs are currently used for drug target recognition in protein-drug interaction networks as well as cancer gene discovery and more. Important aspects whose practical relevance is often underestimated are comprehensibility, interpretability, and explainability. In this work, we present a graph-based deep learning framework for disease subnetwork detection via explainable GNNs. In our framework, each patient is represented by the topology of a protein-protein network (PPI), and the nodes are enriched by molecular multimodal data, such as gene expression and DNA methylation. Therefore, our novel modification of the GNNexplainer for model-wide explanations can detect potential disease subnetworks, which is of high practical relevance. The proposed methods are implemented in the GNN-SubNet Python program, which we have made freely available on our GitHub for the international research community (https://github.com/pievos101/GNN-SubNet).

Microrheology to Understand the Viscosity Behavior of a Sophorolipid Biosurfactant

The microstructure of the aqueous solutions of purified acidic Sophorolipid (SL) has previously been studied using highly sophisticated methods such as SANS and Cryo-TEM. We were interested in whether (a) the main findings also apply to commercially available SL (which is a mixture of acidic and lactonic SL) and (b) more readily available methods such as DLS can be used to gain insight into the molecular aggregation of SL. Our work was motivated by the increasing interest in biosurfactants for applications in personal and household care. Moreover, the origin behind the more or less lack of rheological response to changes in pH is of practical relevance, as it is somewhat unusual for a carboxylate-group containing surfactant. By using DLS microrheology, we could elucidate the aggregation structure and dynamics of the surfactant on a microscopic scale. Surprisingly, the different degrees of protonation only impacted the microscopic properties such as exchange kinetics and the plateau values of the storage moduli.

CRYOGENIC GREASE LUBRICANTS FOR ROCKET AND SPACE TECHNOLOGY

The paper represents requirements for cryogenic grease lubricants used in the rocket and space technology. Data on lubricants based on perfluoropolyether liquid FEN is provided. New lubricant testing methods enabling to analyze their chemical composition and low-temperature characteristics are proposed. Quoted are investigation results for the equivalents of previously used cryogenic lubricants NIKA, NIRA and «Ametist». A practical relevance of the paper has been proven by introduction of low-temperature lubricants «Sever» currently being used in rocket and space items, as well аs successful application of the developed lubricant incoming inspection procedures at RSC Energia. Key words: low-temperature lubricant, perfluoropolyether liquid FEN, viscosity, solidification temperature, IR spectrum, potential evaporation.

Instrumento para mensuração do conhecimento tático declarativo no basquetebol feminino: validade de conteúdo da versão preliminar (Instrument for measurement of declarative tactical knowledge in women's basketball: validity of preliminary version content

O objetivo do estudo foi verificar as evidências de validade de conteúdo de cenas de vídeo extraídas de partidas oficiais de basquetebol para a construção da versão preliminar do Instrumento para Mensurar o Conhecimento Tático Declarativo para Basquetebol Feminino (IMCTD-BF). Editou-se, para o processo de validação de conteúdo, 70 cenas extraídas de jogos entre equipes que disputaram o LBF (Liga de Basquete Feminina). As cenas passaram pela análise de cinco peritos, treinadores (as) de basquetebol, nos quesitos: clareza da imagem, pertinência prática e representatividade do item. Calculou-se o Coeficiente de Validade do Conteúdo (CVC) para verificar a concordância entre os peritos, sendo o ponto de corte .80. Encontrou-se valores de CVC de .956 para clareza da imagem, .997 para pertinência prática e .960 para representatividade do item. Os procedimentos de validade ecológica e a concordância entre peritos, referentes à tomada de decisão e ao tipo de ataque, resultaram na seleção de 38 cenas. Conclui-se que foi possível construir a versão preliminar do IMCTD-BF a partir de 38 cenas que apresentaram evidências de validade de conteúdo. Abstract. The aim of the study was to verify the evidence of content validity of video scenes extracted from official basketball matches for the construction of the preliminary version of the Instrument to Measure Declarative Tactical Knowledge for Women's Basketball (IMCTD-BF). For the content validation process, 70 scenes extracted from games between teams that competed in the LBF (Women's Basketball League) were edited. The scenes were analyzed by five experts, basketball coaches, in the following items: clarity of the image, practical relevance and representativeness of the item. The Content Validity Coefficient (CVC) was calculated to verify the agreement between experts, with a cutoff point of .80. CVC values ​​of .956 were found for image clarity, .997 for practical relevance and .960 for item representativeness. The ecological validity procedures and the agreement between experts, regarding decision-making and the type of attack, resulted in the selection of 38 scenes. It is concluded that it was possible to build the preliminary version of the IMCTD-BF from 38 scenes that presented evidence of content validity. Resumem. El objetivo del estudio fue verificar la evidencia de validez de contenido de escenas de video extraídas de partidos oficiales de baloncesto para la construcción de la versión preliminar del Instrumento para Medir el Conocimiento Táctico Declarativo para el Baloncesto Femenino (IMCTD-BF). Para el proceso de validación de contenido se editaron 70 escenas extraídas de partidos entre equipos que compitieron en la LBF (Liga de Baloncesto Femenino). Las escenas fueron analizadas por cinco expertos, entrenadores de baloncesto, en los ítems: claridad de imagen, relevancia práctica y representatividad del ítem. Se calculó el Coeficiente de Validez de Contenido (CVC) para verificar la concordancia entre expertos, con un punto de corte de .80. Se encontraron valores CVC de .956 para claridad de imagen, .997 para relevancia práctica y .960 para representatividad de ítems. Los procedimientos de validez ecológica y el acuerdo entre expertos, en cuanto a la toma de decisiones y el tipo de atentado, dieron como resultado la selección de 38 escenas. Se concluye que fue posible construir la versión preliminar del IMCTD-BF a partir de 38 escenas que presentaban evidencia de validez de contenido.

Disclaimers

Disclaimers are practical expressions of either renunciation or repudiation of interest of self or other, whether they be drawn up as formal clauses or as improvisations (acceptable substitutes). We use them to regulate social interaction by making certain limitations known to parties. Diversity of situations reflects in diversity of expression and means of enforcing such limitations; therein lies disclaimers’ long and luxuriant history in virtually all cultures. Beyond their origin in the permanent need to amend things of practical relevance, their current popularity relates to the impetus of technological progress in communication and to the emerging forms of authorship and literacy.

Exploring factors that mitigate the continued influence of misinformation

Background The term “continued influence effect” (CIE) refers to the phenomenon that discredited and obsolete information continues to affect behavior and beliefs. The practical relevance of this work is particularly apparent as we confront fake news everyday. Thus, an important question becomes, how can we mitigate the continued influence of misinformation? Decades of research have identified several factors that contribute to the CIE reduction, but few have reported successful elimination. Across three studies, we evaluated the relative contribution of three factors (i.e., targeting the misinformation, providing an alternative explanation, and relative importance of the misinformation content) to the reduction of the CIE. Results Across three studies and two different CIE measures, we found that alternative provision consistently resulted in CIE reduction. Furthermore, under certain conditions, the combination of alternative inclusion and direct targeting of misinformation in the correction statement resulted in successful elimination of the CIE, such that individuals who encountered that type of correction behaved similarly to baseline participants who never encountered the (mis)information. In contrast, under one CIE measure, participants who received correction statements that failed to include those elements referenced the (mis)information as frequently as baseline participants who never encountered a correction. Finally, we delineated several component processes involved in misinformation outdating and found that the extent of outdating success varied as a function of the causality of misinformation. Conclusions The damaging effects of fake news are undeniable, and the negative consequences are exacerbated in the digital age. Our results contribute to our understanding of how fake news persists and how we may begin to mitigate their effects.