Medical Entity Linking (MEL) is a common task in natural language processing, focusing on the normalization of recognized entities from clinical texts using large knowledge bases (KBs). This task presents significant challenges, especially when working with electronic health records that often lack annotated clinical notes, even in languages like English. The difficulty increases in few-shot or zero-shot scenarios, where models must operate with minimal or no training data, a common issue when dealing with less-documented languages such as Spanish. Existing solutions that combine contrastive learning with external sources, like the Unified Medical Language System (UMLS), have shown competitive results. However, most of these methods focus on individual concepts from the KBs, ignoring relationships such as synonymy or hierarchical links between concepts. In this paper, we propose leveraging these relationships to enrich the training triplets used for contrastive learning, improving performance in MEL tasks. Specifically, we fine-tune several BERT-based cross-encoders using enriched triplets on three clinical corpora in Spanish : DisTEMIST, MedProcNER, and SympTEMIST. Our approach addresses the complexity of real-world data, where unseen mentions and concepts are frequent. The results show a notable improvement in lower top- k accuracies, surpassing the state-of-the-art by up to 5.5 percentage points for unseen mentions and by up to 5.9 points for unseen concepts. This improvement reduces the number of candidate concepts required for cross-encoders, enabling more efficient semi-automatic annotation and decreasing human effort. Additionally, our findings underscore the importance of leveraging not only the concept-level information in KBs but also the relationships between those concepts. • Our strategy significantly improves medical entity linking (MEL) performance in zero-shot and few-shot scenarios, achieving up to 5.5% and 5.9% increases in unseen mentions and concepts,respectively. • The enriched approach we propose achieves a top-25 accuracy of around 75% and even higher in the most challenging entity normalization scenarios. • The robustness and efficacy of the triplets maintain stable performance across the three main Spanish medical corpora (DisTEMIST, MedProcNER, SympTEMIST), showing consistent behavior in different medical domains. • We propose a new triplet generation strategy for training cross-encoders using both parent–child and bidirectional hierarchical relationships in knowledge graphs. • The approach enables efficient candidate re-ranking, reducing the manual effort in medical annotation tasks by improving top-k accuracies and minimizing the candidates needed for normalization.

MOTIVATION: Relation extraction (RE) is a crucial process to deal with the amount of text published daily, e.g. to find missing associations in a database. RE is a text mining task for which the state-of-the-art approaches use bidirectional encoders, namely, BERT. However, state-of-the-art performance may be limited by the lack of efficient external knowledge injection approaches, with a larger impact in the biomedical area given the widespread usage and high quality of biomedical ontologies. This knowledge can propel these systems forward by aiding them in predicting more explainable biomedical associations. With this in mind, we developed K-RET, a novel, knowledgeable biomedical RE system that, for the first time, injects knowledge by handling different types of associations, multiple sources and where to apply it, and multi-token entities. RESULTS: We tested K-RET on three independent and open-access corpora (DDI, BC5CDR, and PGR) using four biomedical ontologies handling different entities. K-RET improved state-of-the-art results by 2.68% on average, with the DDI Corpus yielding the most significant boost in performance, from 79.30% to 87.19% in F-measure, representing a P-value of 2.91×10-12. AVAILABILITY AND IMPLEMENTATION: https://github.com/lasigeBioTM/K-RET.

BACKGROUND: Complex diseases such as neurodevelopmental disorders (NDDs) exhibit multiple etiologies. The multi-etiological nature of complex-diseases emerges from distinct but functionally similar group of genes. Different diseases sharing genes of such groups show related clinical outcomes that further restrict our understanding of disease mechanisms, thus, limiting the applications of personalized medicine approaches to complex genetic disorders. RESULTS: Here, we present an interactive and user-friendly application, called DGH-GO. DGH-GO allows biologists to dissect the genetic heterogeneity of complex diseases by stratifying the putative disease-causing genes into clusters that may contribute to distinct disease outcome development. It can also be used to study the shared etiology of complex-diseases. DGH-GO creates a semantic similarity matrix for the input genes by using Gene Ontology (GO). The resultant matrix can be visualized in 2D plots using different dimension reduction methods (T-SNE, Principal component analysis, umap and Principal coordinate analysis). In the next step, clusters of functionally similar genes are identified from genes functional similarities assessed through GO. This is achieved by employing four different clustering methods (K-means, Hierarchical, Fuzzy and PAM). The user may change the clustering parameters and explore their effect on stratification immediately. DGH-GO was applied to genes disrupted by rare genetic variants in Autism Spectrum Disorder (ASD) patients. The analysis confirmed the multi-etiological nature of ASD by identifying four clusters of genes that were enriched for distinct biological mechanisms and clinical outcome. In the second case study, the analysis of genes shared by different NDDs showed that genes causing multiple disorders tend to aggregate in similar clusters, indicating a possible shared etiology. CONCLUSION: DGH-GO is a user-friendly application that allows biologists to study the multi-etiological nature of complex diseases by dissecting their genetic heterogeneity. In summary, functional similarities, dimension reduction and clustering methods, coupled with interactive visualization and control over analysis allows biologists to explore and analyze their datasets without requiring expert knowledge on these methods. The source code of proposed application is available at https://github.com/Muh-Asif/DGH-GO.

The existence of unlinkable (NIL) entities is a major hurdle affecting the performance of Named Entity Linking approaches, and, consequently, the performance of downstream models that depend on them. Existing approaches to deal with NIL entities focus mainly on clustering and prediction and are limited to general entities. However, other domains, such as the biomedical sciences, are also prone to the existence of NIL entities, given the growing nature of scientific literature. We propose NILINKER, a model that includes a candidate retrieval module for biomedical NIL entities and a neural network that leverages the attention mechanism to find the top-k relevant concepts from target Knowledge Bases (MEDIC, CTD-Chemicals, ChEBI, HP, CTD-Anatomy and Gene Ontology-Biological Process) that may partially represent a given NIL entity. We also make available a new evaluation dataset designated by EvaNIL, suitable for training and evaluating models focusing on the NIL entity linking task. This dataset contains 846,165 documents (abstracts and full-text biomedical articles), including 1,071,776 annotations, distributed by six different partitions: EvaNIL-MEDIC, EvaNIL-CTD-Chemicals, EvaNIL-ChEBI, EvaNIL-HP, EvaNIL-CTD-Anatomy and EvaNIL-Gene Ontology-Biological Process. NILINKER was integrated into a graph-based Named Entity Linking model (REEL) and the results of the experiments show that this approach is able to increase the performance of the Named Entity Linking model.

Biomedical Relation Extraction (RE) systems identify and classify relations between biomedical entities to enhance our knowledge of biological and medical processes. Most state-of-the-art systems use deep learning approaches, mainly to target relations between entities of the same type, such as proteins or pharmacological substances. However, these systems are mostly restricted to what they directly identify on the text and ignore specialized domain knowledge bases, such as ontologies, that formalize and integrate biomedical information typically structured as direct acyclic graphs. On the other hand, Knowledge Graph (KG)-based recommendation systems already showed the importance of integrating KGs to add additional features to items. Typical systems have users as people and items that can range from movies to books, which people saw or read and classified according to their satisfaction rate. This work proposes to integrate KGs into biomedical RE through a recommendation model to further improve their range of action. We developed a new RE system, named K-BiOnt, by integrating a baseline state-of-the-art deep biomedical RE system with an existing KG-based recommendation state-of-the-art system. Our results show that adding recommendations from KG-based recommendation improves the system's ability to identify true relations that the baseline deep RE model could not extract from the text. The code supporting this system is available at https://github.com/lasigeBioTM/K-BiOnt.

Recommender systems (RS) have been successfully explored in a vast number of domains, e.g. movies and tv shows, music, or e-commerce. In these domains we have a large number of datasets freely available for testing and evaluating new recommender algorithms. For example, Movielens and Netflix datasets for movies, Spotify for music, and Amazon for e-commerce, which translates into a large number of algorithms applied to these fields. In scientific fields, such as Health and Chemistry, standard and open access datasets with the information about the preferences of the users are scarce. First, it is important to understand the application domain, i.e. "what the recommended item is". Second, who are the end users: researchers, pharmacists, clinicians or policy makers. Third, the availability of data. Thus, if we wish to develop an algorithm for recommending scientific items, we do not have access to datasets with information about the past preferences of a group of users. Given this limitation, we developed a methodology, called LIBRETTI - LIterature Based RecommEndaTion of scienTific Items, whose goal is the creation of datasets, related with scientific fields. These datasets are created based on the major resource of knowledge that Science has: scientific literature. We consider the users as the authors of the publications, the items as the scientific entities (for example chemical compounds or diseases), and the ratings as the number of publications an author wrote about an entity. In this tutorial we will approach state-of-the-art recommender systems in scientific fields, explain what is Named Entity Recognition/Linking (NER/NEL) in research literature, and to demonstrate how to create a dataset for recommending drugs and diseases through research literature related to COVID-19. Our goal is to spread the use of LIBRETTI methodology in order to help in the development of recommender algorithms in scientific fields. These datasets are created based on the major resource of knowledge that Science has: scientific literature. We consider the users as the authors of the publications, the items as the scientific entities (for example chemical compounds or diseases), and the ratings as the number of publications an author wrote about an entity. In this tutorial we will approach state-of-the-art recommender systems in scientific fields, explain what is Named Entity Recognition/Linking (NER/NEL) in research literature, and to demonstrate how to create a dataset for recommending drugs and diseases through research literature related to COVID-19. Our goal is to spread the use of LIBRETTI methodology in order to help in the development of recommender algorithms in scientific fields. More info about the tutorial at https://lasigebiotm.github.io/RecSys.Scifi/.

Frameworks that aid the development of Recommender Systems (RSs) are extremely important, since they reduce their development cost by offering reusable tools, as well as implementations of common strategies and popular models. However, it is still hard to find a framework that also provides full abstraction over data set conversion, support for deep learning-based approaches, extensible models and reproducible evaluations. This work introduces a new framework that not only provides several modules to avoid repetitive development work, but also to assist practitioners with these existing challenges. Our evaluation procedure ensures that RSs developed using this new approach are consistent and extensible, by analysing their predictive performance and certain characteristics of their implementation.

The advent of next-generation sequencing (NGS) machines made DNA sequencing cheaper, but also put pressure on the genomic life-cycle, which includes aligning millions of short DNA sequences, called reads, to a reference genome. On the performance side, efficient algorithms have been developed, and parallelized on public clouds. On the privacy side, since genomic data are utterly sensitive, several cryptographic mechanisms have been proposed to align reads more securely than the former, but with a lower performance. This paper presents DNA-SeAl a novel contribution to improving the privacy × performance product in current genomic workflows. First, building on recent works that argue that genomic data needs to be treated according to a threat-risk analysis, we introduce a multi-level sensitivity classification of genomic variations designed to prevent the amplification of possible privacy attacks. We show that the usage of sensitivity levels reduces future re-identification risks, and that their partitioning helps prevent linkage attacks. Second, after extending this classification to reads, we show how to align and store reads using different security levels. To do so, DNA-SeAl extends a recent reads filter to classify unaligned reads into sensitivity levels, and adapts existing alignment algorithms to the reads sensitivity. We show that using DNA-SeAl allows high performance gains whilst enforcing high privacy levels in hybrid cloud environments.

BACKGROUND: Given the increasing amount of biomedical resources that are being annotated with concepts from more than one ontology and covering multiple domains of knowledge, it is important to devise mechanisms to compare these resources that take into account the various domains of annotation. For example, metabolic pathways are annotated with their enzymes and their metabolites, and thus similarity measures should compare them with respect to both of those domains simultaneously. RESULTS: In this paper, we propose two approaches to lift existing single-ontology semantic similarity measures into multi-domain measures. The aggregative approach compares domains independently and averages the various similarity values into a final score. The integrative approach integrates all the relevant ontologies into a single one, calculating similarity in the resulting multi-domain ontology using the single-ontology measure. CONCLUSIONS: We evaluated the two approaches in a multidisciplinary epidemiology dataset by evaluating the capacity of the similarity measures to predict new annotations based on the existing ones. The results show a promising increase in performance of the multi-domain measures over the single-ontology ones in the vast majority of the cases. These results show that multi-domain measures outperform single-domain ones, and should be considered by the community as a starting point to study more efficient multi-domain semantic similarity measures.

BACKGROUND: Biomedical literature concerns a wide range of concepts, requiring controlled vocabularies to maintain a consistent terminology across different research groups. However, as new concepts are introduced, biomedical literature is prone to ambiguity, specifically in fields that are advancing more rapidly, for example, drug design and development. Entity linking is a text mining task that aims at linking entities mentioned in the literature to concepts in a knowledge base. For example, entity linking can help finding all documents that mention the same concept and improve relation extraction methods. Existing approaches focus on the local similarity of each entity and the global coherence of all entities in a document, but do not take into account the semantics of the domain. RESULTS: We propose a method, PPR-SSM, to link entities found in documents to concepts from domain-specific ontologies. Our method is based on Personalized PageRank (PPR), using the relations of the ontology to generate a graph of candidate concepts for the mentioned entities. We demonstrate how the knowledge encoded in a domain-specific ontology can be used to calculate the coherence of a set of candidate concepts, improving the accuracy of entity linking. Furthermore, we explore weighting the edges between candidate concepts using semantic similarity measures (SSM). We show how PPR-SSM can be used to effectively link named entities to biomedical ontologies, namely chemical compounds, phenotypes, and gene-product localization and processes. CONCLUSIONS: We demonstrated that PPR-SSM outperforms state-of-the-art entity linking methods in four distinct gold standards, by taking advantage of the semantic information contained in ontologies. Moreover, PPR-SSM is a graph-based method that does not require training data. Our method improved the entity linking accuracy of chemical compounds by 0.1385 when compared to a method that does not use SSMs.

BACKGROUND: Recent studies have proposed deep learning techniques, namely recurrent neural networks, to improve biomedical text mining tasks. However, these techniques rarely take advantage of existing domain-specific resources, such as ontologies. In Life and Health Sciences there is a vast and valuable set of such resources publicly available, which are continuously being updated. Biomedical ontologies are nowadays a mainstream approach to formalize existing knowledge about entities, such as genes, chemicals, phenotypes, and disorders. These resources contain supplementary information that may not be yet encoded in training data, particularly in domains with limited labeled data. RESULTS: We propose a new model to detect and classify relations in text, BO-LSTM, that takes advantage of domain-specific ontologies, by representing each entity as the sequence of its ancestors in the ontology. We implemented BO-LSTM as a recurrent neural network with long short-term memory units and using open biomedical ontologies, specifically Chemical Entities of Biological Interest (ChEBI), Human Phenotype, and Gene Ontology. We assessed the performance of BO-LSTM with drug-drug interactions mentioned in a publicly available corpus from an international challenge, composed of 792 drug descriptions and 233 scientific abstracts. By using the domain-specific ontology in addition to word embeddings and WordNet, BO-LSTM improved the F1-score of both the detection and classification of drug-drug interactions, particularly in a document set with a limited number of annotations. We adapted an existing DDI extraction model with our ontology-based method, obtaining a higher F1 score than the original model. Furthermore, we developed and made available a corpus of 228 abstracts annotated with relations between genes and phenotypes, and demonstrated how BO-LSTM can be applied to other types of relations. CONCLUSIONS: Our findings demonstrate that besides the high performance of current deep learning techniques, domain-specific ontologies can still be useful to mitigate the lack of labeled data.

Sequencing thousands of human genomes has enabled breakthroughs in many areas, among them precision medicine, the study of rare diseases, and forensics. However, mass collection of such sensitive data entails enormous risks if not protected to the highest standards. In this article, we follow the position and argue that post-alignment privacy is not enough and that data should be automatically protected as early as possible in the genomics workflow, ideally immediately after the data is produced. We show that a previous approach for filtering short reads cannot extend to long reads and present a novel filtering approach that classifies raw genomic data (i.e., whose location and content is not yet determined) into privacy-sensitive (i.e., more affected by a successful privacy attack) and non-privacy-sensitive information. Such a classification allows the fine-grained and automated adjustment of protective measures to mitigate the possible consequences of exposure, in particular when relying on public clouds. We present the first filter that can be indistinctly applied to reads of any length, i.e., making it usable with any recent or future sequencing technologies. The filter is accurate, in the sense that it detects all known sensitive nucleotides except those located in highly variable regions (less than 10 nucleotides remain undetected per genome instead of 100,000 in previous works). It has far less false positives than previously known methods (10% instead of 60%) and can detect sensitive nucleotides despite sequencing errors (86% detected instead of 56% with 2% of mutations). Finally, practical experiments demonstrate high performance, both in terms of throughput and memory consumption.

Semantic web technologies offer an approach to data integration and sharing, even for resources developed independently or broadly distributed across the web. This approach is particularly suitable for scientific domains that profit from large amounts of data that reside in the public domain and that have to be exploited in combination. Translational medicine is such a domain, which in addition has to integrate private data from the clinical domain with proprietary data from the pharmaceutical domain. In this survey, we present the results of our analysis of translational medicine solutions that follow a semantic web approach. We assessed these solutions in terms of their target medical use case; the resources covered to achieve their objectives; and their use of existing semantic web resources for the purposes of data sharing, data interoperability and knowledge discovery. The semantic web technologies seem to fulfill their role in facilitating the integration and exploration of data from disparate sources, but it is also clear that simply using them is not enough. It is fundamental to reuse resources, to define mappings between resources, to share data and knowledge. All these aspects allow the instantiation of translational medicine at the semantic web-scale, thus resulting in a network of solutions that can share resources for a faster transfer of new scientific results into the clinical practice. The envisioned network of translational medicine solutions is on its way, but it still requires resolving the challenges of sharing protected data and of integrating semantic-driven technologies into the clinical practice.

Web archives already hold together more than 534 billion files and this number continues to grow as new initiatives arise. Searching on all versions of these files acquired throughout time is challenging, since users expect as fast and precise answers from web archives as the ones provided by current web search engines. This work studies, for the first time, how to improve the search effectiveness of web archives, including the creation of novel temporal features that explore the correlation found between web document persistence and relevance. The persistence was analyzed over 14 years of web snapshots. Additionally, we propose a temporal-dependent ranking framework that exploits the variance of web characteristics over time influencing ranking models. Based on the assumption that closer periods are more likely to hold similar web characteristics, our framework learns multiple models simultaneously, each tuned for a specific period. Experimental results show significant improvements over the search effectiveness of single-models that learn from all data independently of its time. Thus, our approach represents an important step forward on the state-of-the-art IR technology usually employed in web archives.

MOTIVATION: Representing domain knowledge in biology has traditionally been accomplished by creating simple hierarchies of classes with textual annotations. Recently, expressive ontology languages, such as Web Ontology Language, have become more widely adopted, supporting axioms that express logical relationships other than class-subclass, e.g. disjointness. This is improving the coverage and validity of the knowledge contained in biological ontologies. However, current semantic tools still need to adapt to this more expressive information. In this article, we propose a method to integrate disjointness axioms, which are being incorporated in real-world ontologies, such as the Gene Ontology and the chemical entities of biological interest ontology, into semantic similarity, the measure that estimates the closeness in meaning between classes. RESULTS: We present a modification of the measure of shared information content, which extends the base measure to allow the incorporation of disjointness information. To evaluate our approach, we applied it to several randomly selected datasets extracted from the chemical entities of biological interest ontology. In 93.8% of these datasets, our measure performed better than the base measure of shared information content. This supports the idea that semantic similarity is more accurate if it extends beyond the hierarchy of classes of the ontology. CONTACT: [email protected]. SUPPLEMENTARY INFORMATION: Supplementary data are available at Bioinformatics online.

Developing and extending a biomedical ontology is a very demanding task that can never be considered complete given our ever-evolving understanding of the life sciences. Extension in particular can benefit from the automation of some of its steps, thus releasing experts to focus on harder tasks. Here we present a strategy to support the automation of change capturing within ontology extension where the need for new concepts or relations is identified. Our strategy is based on predicting areas of an ontology that will undergo extension in a future version by applying supervised learning over features of previous ontology versions. We used the Gene Ontology as our test bed and obtained encouraging results with average f-measure reaching 0.79 for a subset of biological process terms. Our strategy was also able to outperform state of the art change capturing methods. In addition we have identified several issues concerning prediction of ontology evolution, and have delineated a general framework for ontology extension prediction. Our strategy can be applied to any biomedical ontology with versioning, to help focus either manual or semi-automated extension methods on areas of the ontology that need extension.

With the increasing amount of data made available in the chemical field, there is a strong need for systems capable of comparing and classifying chemical compounds in an efficient and effective way. The best approaches existing today are based on the structure-activity relationship premise, which states that biological activity of a molecule is strongly related to its structural or physicochemical properties. This work presents a novel approach to the automatic classification of chemical compounds by integrating semantic similarity with existing structural comparison methods. Our approach was assessed based on the Matthews Correlation Coefficient for the prediction, and achieved values of 0.810 when used as a prediction of blood-brain barrier permeability, 0.694 for P-glycoprotein substrate, and 0.673 for estrogen receptor binding activity. These results expose a significant improvement over the currently existing methods, whose best performances were 0.628, 0.591, and 0.647 respectively. It was demonstrated that the integration of semantic similarity is a feasible and effective way to improve existing chemical compound classification systems. Among other possible uses, this tool helps the study of the evolution of metabolic pathways, the study of the correlation of metabolic networks with properties of those networks, or the improvement of ontologies that represent chemical information.

In recent years, ontologies have become a mainstream topic in biomedical research. When biological entities are described using a common schema, such as an ontology, they can be compared by means of their annotations. This type of comparison is called semantic similarity, since it assesses the degree of relatedness between two entities by the similarity in meaning of their annotations. The application of semantic similarity to biomedical ontologies is recent; nevertheless, several studies have been published in the last few years describing and evaluating diverse approaches. Semantic similarity has become a valuable tool for validating the results drawn from biomedical studies such as gene clustering, gene expression data analysis, prediction and validation of molecular interactions, and disease gene prioritization. We review semantic similarity measures applied to biomedical ontologies and propose their classification according to the strategies they employ: node-based versus edge-based and pairwise versus groupwise. We also present comparative assessment studies and discuss the implications of their results. We survey the existing implementations of semantic similarity measures, and we describe examples of applications to biomedical research. This will clarify how biomedical researchers can benefit from semantic similarity measures and help them choose the approach most suitable for their studies.Biomedical ontologies are evolving toward increased coverage, formality, and integration, and their use for annotation is increasingly becoming a focus of both effort by biomedical experts and application of automated annotation procedures to create corpora of higher quality and completeness than are currently available. Given that semantic similarity measures are directly dependent on these evolutions, we can expect to see them gaining more relevance and even becoming as essential as sequence similarity is today in biomedical research.

BACKGROUND: Several semantic similarity measures have been applied to gene products annotated with Gene Ontology terms, providing a basis for their functional comparison. However, it is still unclear which is the best approach to semantic similarity in this context, since there is no conclusive evaluation of the various measures. Another issue, is whether electronic annotations should or not be used in semantic similarity calculations. RESULTS: We conducted a systematic evaluation of GO-based semantic similarity measures using the relationship with sequence similarity as a means to quantify their performance, and assessed the influence of electronic annotations by testing the measures in the presence and absence of these annotations. We verified that the relationship between semantic and sequence similarity is not linear, but can be well approximated by a rescaled Normal cumulative distribution function. Given that the majority of the semantic similarity measures capture an identical behaviour, but differ in resolution, we used the latter as the main criterion of evaluation. CONCLUSIONS: This work has provided a basis for the comparison of several semantic similarity measures, and can aid researchers in choosing the most adequate measure for their work. We have found that the hybrid simGIC was the measure with the best overall performance, followed by Resnik's measure using a best-match average combination approach. We have also found that the average and maximum combination approaches are problematic since both are inherently influenced by the number of terms being combined. We suspect that there may be a direct influence of data circularity in the behaviour of the results including electronic annotations, as a result of functional inference from sequence similarity.

Erroneous data can often be found in databases, and detecting it is normally a non-trivial task. For example, To cope with the large amount of biological sequences being produced, a significant number of genes and proteins have been annotated by automated tools. A protein annotation is an association between a protein and a term describing its role. These tools have produced a significant number of misannotations that are now present in biological databases. This paper proposes a new method for automatically scoring associations by comparing them to preexisting curated associations. An association is a pair that links two entities. The score can be used to filter incorrect or uncommon associations.We evaluated the method using the automated protein annotations submitted to BioCreAtIvE, an international evaluation of state-of-the-art text-mining systems in Biology. The method scored each of these annotations and those scored below a certain threshold were discarded. The results have shown a small trade-off in recall for a large improvement in precision. For example, we were able to discard 44.6%, 66.8% and 81% of the misannotations, maintaining 96.9%, 84.2%, and 47.8% of the correct annotations, respectively. Moreover, we were able to outperform each individual submission to BioCreAtIvE by proper adjustment of the threshold.

Many bioinformatics applications would benefit from comparing proteins based on their biological role rather than their sequence. In most biological databases, proteins are already annotated with ontology terms. Previous studies identified a correlation between the sequence similarity and the semantic similarity of proteins. The semantic similarity of proteins was computed from their annotated GO terms. However, proteins sharing a biological role do not necessarily have a similar sequence.This paper introduces our study of the correlation between GO and family similarity. Family similarity overcomes some of the limitations of sequence similarity, thus we obtained a strong correlation between GO and family similarity. Additionally, this paper introduces GraSM, a novel method that uses all the information in the graph structure of the GO, instead of considering it as a hierarchical tree. When calculating the semantic similarity of two concepts, GraSM selects the disjunctive common ancestors rather than only using the most informative common ancestor. GraSM produced a higher family similarity correlation than the original semantic similarity measures.

BACKGROUND: The development of text mining systems that annotate biological entities with their properties using scientific literature is an important recent research topic. These systems need first to recognize the biological entities and properties in the text, and then decide which pairs represent valid annotations. METHODS: This document introduces a novel unsupervised method for recognizing biological properties in unstructured text, involving the evidence content of their names. RESULTS: This document shows the results obtained by the application of our method to BioCreative tasks 2.1 and 2.2, where it identified Gene Ontology annotations and their evidence in a set of articles. CONCLUSION: From the performance obtained in BioCreative, we concluded that an automatic annotation system can effectively use our method to identify biological properties in unstructured text.