Collecting the relevant list of patient phenotypes, known as deep phenotyping, can significantly improve the final diagnosis. As textual clinical reports are the richest source of phenotypes information, their automatic extraction is a critical task. The main challenges of this Information Extraction (IE) task are to identify precisely the text spans related to a phenotype and to link them unequivocally to referenced entities from a source such as the Human Phenotype Ontology (HPO). Recently, Language Models (LMs) have been the most suc-cessful approach for extracting phenotypes from clinical reports. Solutions such as PhenoBERT, relying on BERT or GPT, have shown promising results when applied to datasets built on the hypothesis that most phenotypes are explicitly mentioned in the text. However, this assumption is not always true in medical genetics. Hence, although the LMs carry powerful semantic abilities, their contributions are not clear compared to syntactic string-matching steps that are used within the current pipelines. The goal of this study is to improve phenotype extraction from clinical notes related to genetic diseases. Our contributions are threefold: First, we provide a clear definition of the phenotype extraction task from free text, along with a high-level overview of the involved functions. Second, we conduct an in-depth analysis of PhenoBERT, one of the best existing solutions, to evaluate the proportion of phenotypes predicted with simple string-matching. Third, we demonstrate how utilizing and incorporating large language models (LLMs) for span detection step can improve performance especially with implicit phenotypes. In addition, this experiment revealed that the annotations of existing dataset are not exhaustive, and that LLM can identify relevant spans missed by human labelers.

MOTIVATION: Transcriptional regulation is performed by transcription factors (TF) binding to DNA in context-dependent regulatory regions and determines the activation or inhibition of gene expression. Current methods of transcriptional regulatory circuits inference, based on one or all of TF, regions and genes activity measurements require a large number of samples for ranking the candidate TF-gene regulation relations and rarely predict whether they are activations or inhibitions. We hypothesize that transcriptional regulatory circuits can be inferred from fewer samples by (1) fully integrating information on TF binding, gene expression and regulatory regions accessibility, (2) reducing data complexity and (3) using biology-based likelihood constraints to determine the global consistency between a candidate TF-gene relation and patterns of genes expressions and region activations, as well as qualify regulations as activations or inhibitions. RESULTS: We introduce Regulus, a method which computes TF-gene relations from gene expressions, regulatory region activities and TF binding sites data, together with the genomic locations of all entities. After aggregating gene expressions and region activities into patterns, data are integrated into a RDF (Resource Description Framework) endpoint. A dedicated SPARQL (SPARQL Protocol and RDF Query Language) query retrieves all potential relations between expressed TF and genes involving active regulatory regions. These TF-region-gene relations are then filtered using biological likelihood constraints allowing to qualify them as activation or inhibition. Regulus provides signed relations consistent with public databases and, when applied to biological data, identifies both known and potential new regulators. Regulus is devoted to context-specific transcriptional circuits inference in human settings where samples are scarce and cell populations are closely related, using discretization into patterns and likelihood reasoning to decipher the most robust regulatory relations.

MOTIVATION: Molecular complexes play a major role in the regulation of biological pathways. The Biological Pathway Exchange format (BioPAX) facilitates the integration of data sources describing interactions some of which involving complexes. The BioPAX specification explicitly prevents complexes to have any component that is another complex (unless this component is a black-box complex whose composition is unknown). However, we observed that the well-curated Reactome pathway database contains such recursive complexes of complexes. We propose reproductible and semantically rich SPARQL queries for identifying and fixing invalid complexes in BioPAX databases, and evaluate the consequences of fixing these nonconformities in the Reactome database. RESULTS: For the Homo sapiens version of Reactome, we identify 5833 recursively defined complexes out of the 14 987 complexes (39%). This situation is not specific to the Human dataset, as all tested species of Reactome exhibit between 30% (Plasmodium falciparum) and 40% (Sus scrofa, Bos taurus, Canis familiaris, and Gallus gallus) of recursive complexes. As an additional consequence, the procedure also allows the detection of complex redundancies. Overall, this method improves the conformity and the automated analysis of the graph by repairing the topology of the complexes in the graph. This will allow to apply further reasoning methods on better consistent data. AVAILABILITY AND IMPLEMENTATION: We provide a Jupyter notebook detailing the analysis https://github.com/cjuigne/non_conformities_detection_biopax.

MOTIVATION: Information on protein-protein interactions is collected in numerous primary databases with their own curation process. Several meta-databases aggregate primary databases to provide more exhaustive datasets. In addition to exhaustivity, aggregation contributes to reliability by providing an overview of the various studies and detection methods supporting an interaction. However, interactions listed in different primary databases are partly redundant because some publications reporting protein-protein interactions have been curated by multiple primary databases. Mere aggregation can thus introduce a bias if these redundancies are not identified and eliminated. To overcome this bias, meta-databases rely on the Molecular Interaction ontology that describes interaction detection methods, but they do not fully take advantage of the ontology's rich semantics, which leads to systematically overestimating interaction reproducibility. RESULTS: We propose a precise definition of explicit and implicit redundancy and show that both can be easily detected using Semantic Web technologies. We apply this process to a dataset from the Agile Protein Interactomes DataServer (APID) meta-database and show that while explicit redundancies were detected by the APID aggregation process, about 15% of APID entries are implicitly redundant and should not be taken into account when presenting confidence-related metrics. More than 90% of implicit redundancies result from the aggregation of distinct primary databases, whereas the remaining occurs between entries of a single database. Finally, we build a 'reproducible interactome' with interactions that have been reproduced by multiple methods or publications. The size of the reproducible interactome is drastically impacted by removing redundancies for both yeast (-59%) and human (-56%), and we show that this is largely due to implicit redundancies. AVAILABILITY AND IMPLEMENTATION: Software, data and results are available at https://gitlab.com/nnet56/reproducible-interactome, https://reproducible-interactome.genouest.org/, Zenodo (https://doi.org/10.5281/zenodo.5595037) and NDEx (https://doi.org/10.18119/N94302 and https://doi.org/10.18119/N97S4D). SUPPLEMENTARY INFORMATION: Supplementary data are available at Bioinformatics online.

In life sciences, current standardization and integration efforts are directed towards reference data and knowledge bases. However, original studies results are generally provided in non standardized and specific formats. In addition, the only formalization of analysis pipelines is often limited to textual descriptions in the method sections. Both factors impair the results reproducibility, their maintenance and their reuse for advancing other studies. Semantic Web technologies have proven their efficiency for facilitating the integration and reuse of reference data and knowledge bases. We thus hypothesize that Semantic Web technologies also facilitate reproducibility and reuse of life sciences studies involving pipelines that compute associations between entities according to intermediary relations and dependencies. In order to assess this hypothesis, we considered a case-study in systems biology (http://regulatorycircuits.org), which provides tissue-specific regulatory interaction networks to elucidate perturbations across complex diseases. Our approach consisted in surveying the complete set of provided supplementary files to reveal the underlying structure between the biological entities described in the data. We relied on this structure and used Semantic Web technologies (i) to integrate the Regulatory Circuits data, and (ii) to formalize the analysis pipeline as SPARQL queries. Our result was a 335,429,988 triples dataset on which two SPARQL queries were sufficient to extract each single tissuespecific regulatory network.

SUMMARY: In public health research and more precisely in the reuse of electronic health data, selecting patients, identifying specific events and interpreting results typically requires biomedical knowledge. The queryMed R package aims to facilitate the integration of medical and pharmacological knowledge stored in formats compliant with the Linked Data paradigm (e.g. OWL ontologies and RDF datasets) into the R statistical programming environment. We show how it allowed us to identify all the drugs prescribed for critical limb ischemia (CLI) and also to detect one contraindicated prescription for one patient by linking a medical database of 1003 CLI patients to ontologies. AVAILABILITY AND IMPLEMENTATION: queryMed is readily usable for medical data mappings and enrichment. Sources, R vignettes and test data are available on GitHub (https://github.com/yannrivault/queryMed) and are archived on Zenodo (https://doi.org/10.5281/zenodo.1323481).

Genome-scale metabolic models have become the tool of choice for the global analysis of microorganism metabolism, and their reconstruction has attained high standards of quality and reliability. Improvements in this area have been accompanied by the development of some major platforms and databases, and an explosion of individual bioinformatics methods. Consequently, many recent models result from "à la carte" pipelines, combining the use of platforms, individual tools and biological expertise to enhance the quality of the reconstruction. Although very useful, introducing heterogeneous tools, that hardly interact with each other, causes loss of traceability and reproducibility in the reconstruction process. This represents a real obstacle, especially when considering less studied species whose metabolic reconstruction can greatly benefit from the comparison to good quality models of related organisms. This work proposes an adaptable workspace, AuReMe, for sustainable reconstructions or improvements of genome-scale metabolic models involving personalized pipelines. At each step, relevant information related to the modifications brought to the model by a method is stored. This ensures that the process is reproducible and documented regardless of the combination of tools used. Additionally, the workspace establishes a way to browse metabolic models and their metadata through the automatic generation of ad-hoc local wikis dedicated to monitoring and facilitating the process of reconstruction. AuReMe supports exploration and semantic query based on RDF databases. We illustrate how this workspace allowed handling, in an integrated way, the metabolic reconstructions of non-model organisms such as an extremophile bacterium or eukaryote algae. Among relevant applications, the latter reconstruction led to putative evolutionary insights of a metabolic pathway.

OBJECTIVE: Biomedical research increasingly relies on the integration of information from multiple heterogeneous data sources. Despite the fact that structural and terminological aspects of interoperability are interdependent and rely on a common set of requirements, current efforts typically address them in isolation. We propose a unified ontology-based knowledge framework to facilitate interoperability between heterogeneous sources, and investigate if using the LexEVS terminology server is a viable implementation method. MATERIALS AND METHODS: We developed a framework based on an ontology, the general information model (GIM), to unify structural models and terminologies, together with relevant mapping sets. This allowed a uniform access to these resources within LexEVS to facilitate interoperability by various components and data sources from implementing architectures. RESULTS: Our unified framework has been tested in the context of the EU Framework Program 7 TRANSFoRm project, where it was used to achieve data integration in a retrospective diabetes cohort study. The GIM was successfully instantiated in TRANSFoRm as the clinical data integration model, and necessary mappings were created to support effective information retrieval for software tools in the project. CONCLUSIONS: We present a novel, unifying approach to address interoperability challenges in heterogeneous data sources, by representing structural and semantic models in one framework. Systems using this architecture can rely solely on the GIM that abstracts over both the structure and coding. Information models, terminologies and mappings are all stored in LexEVS and can be accessed in a uniform manner (implementing the HL7 CTS2 service functional model). The system is flexible and should reduce the effort needed from data sources personnel for implementing and managing the integration.

We propose an algorithm for checking the semantic compatibility of Web service parameters and for suggesting compatible parameters pairings between two Web services semi-automatically. This algorithm supports semi-automatic Web service composition in workflows where the order of the Web service is known. In our use case, we used the OWL-S semantic description of Web service. We automatically generated a Taverna-compatible Xscufl file whenever possible.