A deep-learning approach to predict bleeding risk over time in patients on extended anticoagulation therapy.

BACKGROUND: Thus far, all the clinical models developed to predict major bleeding in patients on extended anticoagulation therapy use the baseline predictors to stratify patients into different risk groups. Therefore, these models do not account for the clinical changes and events that occur after the baseline visit, which can modify risk of bleeding. However, it is difficult to develop predictive models from the routine follow-up clinical interviews, which are irregular sequences of multivariate time series data. OBJECTIVES: To demonstrate that deep learning can incorporate patient time series follow-up data to improve prediction of major bleeding. METHODS: We used the baseline and follow-up data that were collected over 8 years in a longitudinal cohort study of 2542 patients, of whom 118 had major bleeding. Four supervised neural network-based machine-learning models were trained on the baseline, follow-up, or both datasets using 70% of the data. The performance of these models was evaluated, along with modified versions of 6 previously developed clinical models, on the remaining 30% of the data. RESULTS: An ensemble of feedforward and recurrent neural networks that used the baseline and follow-up data was the best-performing model, achieving a sensitivity and a specificity of 61% and 82%, respectively, in identifying major bleeding, and it outperformed the previously developed clinical models in terms of area under the receiver operating characteristic curve (82%) and area under the precision-recall curve (14%). CONCLUSION: Time series follow-up data can improve major bleeding prediction in patients on extended anticoagulation therapy.

Computational Identification of Human Biological Processes and Protein Sequence Motifs Putatively Targeted by SARS-CoV-2 Proteins Using Protein-Protein Interaction Networks.

While the COVID-19 pandemic is causing important loss of life, knowledge of the effects of the causative SARS-CoV-2 virus on human cells is currently limited. Investigating protein-protein interactions (PPIs) between viral and host proteins can provide a better understanding of the mechanisms exploited by the virus and enable the identification of potential drug targets. We therefore performed an in-depth computational analysis of the interactome of SARS-CoV-2 and human proteins in infected HEK 293 cells published by Gordon et al. (Nature2020, 583, 459-468) to reveal processes that are potentially affected by the virus and putative protein binding sites. Specifically, we performed a set of network-based functional and sequence motif enrichment analyses on SARS-CoV-2-interacting human proteins and on PPI networks generated by supplementing viral-host PPIs with known interactions. Using a novel implementation of our GoNet algorithm, we identified 329 Gene Ontology terms for which the SARS-CoV-2-interacting human proteins are significantly clustered in PPI networks. Furthermore, we present a novel protein sequence motif discovery approach, LESMoN-Pro, that identified 9 amino acid motifs for which the associated proteins are clustered in PPI networks. Together, these results provide insights into the processes and sequence motifs that are putatively implicated in SARS-CoV-2 infection and could lead to potential therapeutic targets.