SUMMARY: Relation extraction (RE) from large text collections is an important tool for database curation, pathway reconstruction, or functional omics data analysis. In practice, RE often is part of a complex data analysis pipeline requiring specific adaptations like restricting the types of relations or the set of proteins to be considered. However, current systems are either non-programmable web sites or research code with fixed functionality. We present PEDL+, a user-friendly tool for extracting protein-protein and protein-chemical associations from PubMed articles. PEDL+ combines state-of-the-art NLP technology with adaptable ranking and filtering options and can easily be integrated into analysis pipelines. We evaluated PEDL+ in two pathway curation projects and found that 59% to 80% of its extractions were helpful. AVAILABILITY AND IMPLEMENTATION: PEDL+ is freely available at https://github.com/leonweber/pedl.

MOTIVATION: A significant portion of molecular biology investigates signalling pathways and thus depends on an up-to-date and complete resource of functional protein-protein associations (PPAs) that constitute such pathways. Despite extensive curation efforts, major pathway databases are still notoriously incomplete. Relation extraction can help to gather such pathway information from biomedical publications. Current methods for extracting PPAs typically rely exclusively on rare manually labelled data which severely limits their performance. RESULTS: We propose PPA Extraction with Deep Language (PEDL), a method for predicting PPAs from text that combines deep language models and distant supervision. Due to the reliance on distant supervision, PEDL has access to an order of magnitude more training data than methods solely relying on manually labelled annotations. We introduce three different datasets for PPA prediction and evaluate PEDL for the two subtasks of predicting PPAs between two proteins, as well as identifying the text spans stating the PPA. We compared PEDL with a recently published state-of-the-art model and found that on average PEDL performs better in both tasks on all three datasets. An expert evaluation demonstrates that PEDL can be used to predict PPAs that are missing from major pathway databases and that it correctly identifies the text spans supporting the PPA. AVAILABILITY AND IMPLEMENTATION: PEDL is freely available at https://github.com/leonweber/pedl. The repository also includes scripts to generate the used datasets and to reproduce the experiments from this article. SUPPLEMENTARY INFORMATION: Supplementary data are available at Bioinformatics online.

The transcription factors NF-κB and p53 are key regulators in the genotoxic stress response and are critical for tumor development. Although there is ample evidence for interactions between both networks, a comprehensive understanding of the crosstalk is lacking. Here, we developed a systematic approach to identify potential interactions between the pathways. We perturbed NF-κB signaling by inhibiting IKK2, a critical regulator of NF-κB activity, and monitored the altered response of p53 to genotoxic stress using single cell time lapse microscopy. Fitting subpopulation-specific computational p53 models to this time-resolved single cell data allowed to reproduce in a quantitative manner signaling dynamics and cellular heterogeneity for the unperturbed and perturbed conditions. The approach enabled us to untangle the integrated effects of IKK/ NF-κB perturbation on p53 dynamics and thereby derive potential interactions between both networks. Intriguingly, we find that a simultaneous perturbation of multiple processes is necessary to explain the observed changes in the p53 response. Specifically, we show interference with the activation and degradation of p53 as well as the degradation of Mdm2. Our results highlight the importance of the crosstalk and its potential implications in p53-dependent cellular functions.

Oscillations occur in a wide variety of cellular processes, for example in calcium and p53 signaling responses, in metabolic pathways or within gene-regulatory networks, e.g. the circadian system. Since it is of central importance to understand the influence of perturbations on the dynamics of these systems a number of experimental and theoretical studies have examined their robustness. The period of circadian oscillations has been found to be very robust and to provide reliable timing. For intracellular calcium oscillations the period has been shown to be very sensitive and to allow for frequency-encoded signaling. We here apply a comprehensive computational approach to study the robustness of period and amplitude of oscillatory systems. We employ different prototype oscillator models and a large number of parameter sets obtained by random sampling. This framework is used to examine the effect of three design principles on the sensitivities towards perturbations of the kinetic parameters. We find that a prototype oscillator with negative feedback has lower period sensitivities than a prototype oscillator relying on positive feedback, but on average higher amplitude sensitivities. For both oscillator types, the use of Michaelis-Menten instead of mass action kinetics in all degradation and conversion reactions leads to an increase in period as well as amplitude sensitivities. We observe moderate changes in sensitivities if replacing mass conversion reactions by purely regulatory reactions. These insights are validated for a set of established models of various cellular rhythms. Overall, our work highlights the importance of reaction kinetics and feedback type for the variability of period and amplitude and therefore for the establishment of predictive models.