MOTIVATION: Hundreds of resources assign identifiers to biomedical concepts including genes, small molecules, biological processes, diseases, and cell types. Often, these resources overlap by assigning identifiers to the same or related concepts. This creates a data interoperability bottleneck, as integrating data sets and knowledge bases that use identifiers for the same concepts from different resources requires such identifiers to be mapped to each other. However, available mappings are incomplete and fragmented across individual resources, motivating their large-scale integration. RESULTS: We developed SeMRA, a software tool that integrates mappings from multiple sources into a graph data structure. Using graph algorithms, it infers missing mappings implied by available ones while keeping track of provenance and confidence. This allows connecting identifier spaces between which direct mapping was previously not possible. SeMRA implements a customizable workflow that takes a declarative specification as input describing sources to integrate with additional configuration parameters. We used SeMRA to produce the SeMRA Raw Mappings Database, an aggregation of 43.4 million mappings from 127 sources that jointly cover identifiers from 445 ontologies and databases. We also describe benchmarks on specific use cases such as integrating mappings between resources cataloging diseases and cell types. AVAILABILITY AND IMPLEMENTATION: The code is available under the MIT license at https://github.com/biopragmatics/semra. The SeMRA Raw Mappings Database assembled by SeMRA is available at https://doi.org/10.5281/zenodo.11082038.

SUMMARY: We introduce Eliater, a Python package for estimating the effect of perturbation of an upstream molecule on a downstream molecule in a biomolecular network. The estimation takes as input a biomolecular network, observational biomolecular data, and a perturbation of interest, and outputs an estimated quantitative effect of the perturbation. We showcase the functionalities of Eliater in a case study of Escherichia coli transcriptional regulatory network. AVAILABILITY AND IMPLEMENTATION: The code, the documentation, and several case studies are available open source at https://github.com/y0-causal-inference/eliater.

MOTIVATION: Biomedical identifier resources (such as ontologies, taxonomies, and controlled vocabularies) commonly overlap in scope and contain equivalent entries under different identifiers. Maintaining mappings between these entries is crucial for interoperability and the integration of data and knowledge. However, there are substantial gaps in available mappings motivating their semi-automated curation. RESULTS: Biomappings implements a curation workflow for missing mappings which combines automated prediction with human-in-the-loop curation. It supports multiple prediction approaches and provides a web-based user interface for reviewing predicted mappings for correctness, combined with automated consistency checking. Predicted and curated mappings are made available in public, version-controlled resource files on GitHub. Biomappings currently makes available 9274 curated mappings and 40 691 predicted ones, providing previously missing mappings between widely used identifier resources covering small molecules, cell lines, diseases, and other concepts. We demonstrate the value of Biomappings on case studies involving predicting and curating missing mappings among cancer cell lines as well as small molecules tested in clinical trials. We also present how previously missing mappings curated using Biomappings were contributed back to multiple widely used community ontologies. AVAILABILITY AND IMPLEMENTATION: The data and code are available under the CC0 and MIT licenses at https://github.com/biopragmatics/biomappings.

MOTIVATION: The investigation of sets of genes using biological pathways is a common task for researchers and is supported by a wide variety of software tools. This type of analysis generates hypotheses about the biological processes that are active or modulated in a specific experimental context. RESULTS: The Network Data Exchange Integrated Query (NDEx IQuery) is a new tool for network and pathway-based gene set interpretation that complements or extends existing resources. It combines novel sources of pathways, integration with Cytoscape, and the ability to store and share analysis results. The NDEx IQuery web application performs multiple gene set analyses based on diverse pathways and networks stored in NDEx. These include curated pathways from WikiPathways and SIGNOR, published pathway figures from the last 27 years, machine-assembled networks using the INDRA system, and the new NCI-PID v2.0, an updated version of the popular NCI Pathway Interaction Database. NDEx IQuery's integration with MSigDB and cBioPortal now provides pathway analysis in the context of these two resources. AVAILABILITY AND IMPLEMENTATION: NDEx IQuery is available at https://www.ndexbio.org/iquery and is implemented in Javascript and Java.

In this paper, we introduce ChemicalX, a PyTorch-based deep learning library designed for providing a range of state of the art models to solve the drug pair scoring task. The primary objective of the library is to make deep drug pair scoring models accessible to machine learning researchers and practitioners in a streamlined framework. The design of ChemicalX reuses existing high level model training utilities, geometric deep learning, and deep chemistry layers from the PyTorch ecosystem. Our system provides neural network layers, custom pair scoring architectures, data loaders, and batch iterators for end users. We showcase these features with example code snippets and case studies to highlight the characteristics of ChemicalX. A range of experiments on real world drug-drug interaction, polypharmacy side effect, and combination synergy prediction tasks demonstrate that the models available in ChemicalX are effective at solving the pair scoring task. Finally, we show that ChemicalX could be used to train and score machine learning models on large drug pair datasets with hundreds of thousands of compounds on commodity hardware.

MOTIVATION: The majority of biomedical knowledge is stored in structured databases or as unstructured text in scientific publications. This vast amount of information has led to numerous machine learning-based biological applications using either text through natural language processing (NLP) or structured data through knowledge graph embedding models. However, representations based on a single modality are inherently limited. RESULTS: To generate better representations of biological knowledge, we propose STonKGs, a Sophisticated Transformer trained on biomedical text and Knowledge Graphs (KGs). This multimodal Transformer uses combined input sequences of structured information from KGs and unstructured text data from biomedical literature to learn joint representations in a shared embedding space. First, we pre-trained STonKGs on a knowledge base assembled by the Integrated Network and Dynamical Reasoning Assembler consisting of millions of text-triple pairs extracted from biomedical literature by multiple NLP systems. Then, we benchmarked STonKGs against three baseline models trained on either one of the modalities (i.e. text or KG) across eight different classification tasks, each corresponding to a different biological application. Our results demonstrate that STonKGs outperforms both baselines, especially on the more challenging tasks with respect to the number of classes, improving upon the F1-score of the best baseline by up to 0.084 (i.e. from 0.881 to 0.965). Finally, our pre-trained model as well as the model architecture can be adapted to various other transfer learning applications. AVAILABILITY AND IMPLEMENTATION: We make the source code and the Python package of STonKGs available at GitHub (https://github.com/stonkgs/stonkgs) and PyPI (https://pypi.org/project/stonkgs/). The pre-trained STonKGs models and the task-specific classification models are respectively available at https://huggingface.co/stonkgs/stonkgs-150k and https://zenodo.org/communities/stonkgs. SUPPLEMENTARY INFORMATION: Supplementary data are available at Bioinformatics online.

Biological systems are acknowledged to be robust to perturbations but a rigorous understanding of this has been elusive. In a mathematical model, perturbations often exert their effect through parameters, so sizes and shapes of parametric regions offer an integrated global estimate of robustness. Here, we explore this "parameter geography" for bistability in post-translational modification (PTM) systems. We use the previously developed "linear framework" for timescale separation to describe the steady-states of a two-site PTM system as the solutions of two polynomial equations in two variables, with eight non-dimensional parameters. Importantly, this approach allows us to accommodate enzyme mechanisms of arbitrary complexity beyond the conventional Michaelis-Menten scheme, which unrealistically forbids product rebinding. We further use the numerical algebraic geometry tools Bertini, Paramotopy, and alphaCertified to statistically assess the solutions to these equations at ∼109 parameter points in total. Subject to sampling limitations, we find no bistability when substrate amount is below a threshold relative to enzyme amounts. As substrate increases, the bistable region acquires 8-dimensional volume which increases in an apparently monotonic and sigmoidal manner towards saturation. The region remains connected but not convex, albeit with a high visibility ratio. Surprisingly, the saturating bistable region occupies a much smaller proportion of the sampling domain under mechanistic assumptions more realistic than the Michaelis-Menten scheme. We find that bistability is compromised by product rebinding and that unrealistic assumptions on enzyme mechanisms have obscured its parametric rarity. The apparent monotonic increase in volume of the bistable region remains perplexing because the region itself does not grow monotonically: parameter points can move back and forth between monostability and bistability. We suggest mathematical conjectures and questions arising from these findings. Advances in theory and software now permit insights into parameter geography to be uncovered by high-dimensional, data-centric analysis.

SUMMARY: INDRA-IPM (Interactive Pathway Map) is a web-based pathway map modeling tool that combines natural language processing with automated model assembly and visualization. INDRA-IPM contextualizes models with expression data and exports them to standard formats. AVAILABILITY AND IMPLEMENTATION: INDRA-IPM is available at: http://pathwaymap.indra.bio. Source code is available at http://github.com/sorgerlab/indra_pathway_map. The underlying web service API is available at http://api.indra.bio:8000. SUPPLEMENTARY INFORMATION: Supplementary data are available at Bioinformatics online.

BACKGROUND: For automated reading of scientific publications to extract useful information about molecular mechanisms it is critical that genes, proteins and other entities be correctly associated with uniform identifiers, a process known as named entity linking or "grounding." Correct grounding is essential for resolving relationships among mined information, curated interaction databases, and biological datasets. The accuracy of this process is largely dependent on the availability of machine-readable resources associating synonyms and abbreviations commonly found in biomedical literature with uniform identifiers. RESULTS: In a task involving automated reading of ∼215,000 articles using the REACH event extraction software we found that grounding was disproportionately inaccurate for multi-protein families (e.g., "AKT") and complexes with multiple subunits (e.g."NF- κB"). To address this problem we constructed FamPlex, a manually curated resource defining protein families and complexes as they are commonly encountered in biomedical text. In FamPlex the gene-level constituents of families and complexes are defined in a flexible format allowing for multi-level, hierarchical membership. To create FamPlex, text strings corresponding to entities were identified empirically from literature and linked manually to uniform identifiers; these identifiers were also mapped to equivalent entries in multiple related databases. FamPlex also includes curated prefix and suffix patterns that improve named entity recognition and event extraction. Evaluation of REACH extractions on a test corpus of ∼54,000 articles showed that FamPlex significantly increased grounding accuracy for families and complexes (from 15 to 71%). The hierarchical organization of entities in FamPlex also made it possible to integrate otherwise unconnected mechanistic information across families, subfamilies, and individual proteins. Applications of FamPlex to the TRIPS/DRUM reading system and the Biocreative VI Bioentity Normalization Task dataset demonstrated the utility of FamPlex in other settings. CONCLUSION: FamPlex is an effective resource for improving named entity recognition, grounding, and relationship resolution in automated reading of biomedical text. The content in FamPlex is available in both tabular and Open Biomedical Ontology formats at https://github.com/sorgerlab/famplex under the Creative Commons CC0 license and has been integrated into the TRIPS/DRUM and REACH reading systems.

We consider the problem of parameter synthesis for black-box systems whose operations are jointly influenced by a set of “tunable parameters” under the control of designers, and a set of uncontrollable stochastic parameters. The goal is to find values of the tunable parameters that ensure the satisfaction of given performance requirements with a high probability. Such problems are common in robust system design, including feedback controllers, biomedical devices, and many others. These can be naturally cast as chance-constrained optimization problems, which however, are hard to solve precisely. We present a simulation-based approach that provides a piecewise approximation of a certain quantile function for the responses of interest. Using the piecewise approximations as objective functions, a collection of local optima are estimated, from which a global search based on simulated annealing is performed. The search yields tunable parameter values at which the performance requirements are satisfied with a high probability, despite variations in the stochastic parameters. Our approach is applied to three benchmarks: an insulin infusion pump model for type-1 diabetic patients, a robust flight control problem for fixed-wing aircrafts, and an ODE-based apoptosis model from system biology.