Florian Leitner

Overview of the protein-protein interaction annotation extraction task of BioCreative II

Background:The biomedical literature is the primary information source for manual protein-protein interaction annotations. Text-mining systems have been implemented to extract binary protein interactions from articles, but a comprehensive comparison between the different techniques as well as with manual curation was missing.Results:We designed a community challenge, the BioCreative II protein-protein interaction (PPI) task, based on the main steps of a manual protein interaction annotation workflow. It was structured into four distinct subtasks related to: (a) detection of protein interaction-relevant articles; (b) extraction and normalization of protein interaction pairs; (c) retrieval of the interaction detection methods used; and (d) retrieval of actual text passages that provide evidence for protein interactions. A total of 26 teams submitted runs for at least one of the proposed subtasks. In the interaction article detection subtask, the top scoring team reached an F-score of 0.78. In the interaction pair extraction and mapping to SwissProt, a precision of 0.37 (with recall of 0.33) was obtained. For associating articles with an experimental interaction detection method, an F-score of 0.65 was achieved. As for the retrieval of the PPI passages best summarizing a given protein interaction in full-text articles, 19% of the submissions returned by one of the runs corresponded to curator-selected sentences. Curators extracted only the passages that best summarized a given interaction, implying that many of the automatically extracted ones could contain interaction information but did not correspond to the most informative sentences.Conclusion:The BioCreative II PPI task is the first attempt to compare the performance of text-mining tools specific for each of the basic steps of the PPI extraction pipeline. The challenges identified range from problems in full-text format conversion of articles to difficulties in detecting interactor protein pairs and then linking them to their database records. Some limitations were also encountered when using a single (and possibly incomplete) reference database for protein normalization or when limiting search for interactor proteins to co-occurrence within a single sentence, when a mention might span neighboring sentences. Finally, distinguishing between novel, experimentally verified interactions (annotation relevant) and previously known interactions adds additional complexity to these tasks.

The Protein-Protein Interaction tasks of BioCreative III: classification/ranking of articles and linking bio-ontology concepts to full text

BackgroundDetermining usefulness of biomedical text mining systems requires realistic task definition and data selection criteria without artificial constraints, measuring performance aspects that go beyond traditional metrics. The BioCreative III Protein-Protein Interaction (PPI) tasks were motivated by such considerations, trying to address aspects including how the end user would oversee the generated output, for instance by providing ranked results, textual evidence for human interpretation or measuring time savings by using automated systems. Detecting articles describing complex biological events like PPIs was addressed in the Article Classification Task (ACT), where participants were asked to implement tools for detecting PPI-describing abstracts. Therefore the BCIII-ACT corpus was provided, which includes a training, development and test set of over 12,000 PPI relevant and non-relevant PubMed abstracts labeled manually by domain experts and recording also the human classification times. The Interaction Method Task (IMT) went beyond abstracts and required mining for associations between more than 3,500 full text articles and interaction detection method ontology concepts that had been applied to detect the PPIs reported in them.ResultsA total of 11 teams participated in at least one of the two PPI tasks (10 in ACT and 8 in the IMT) and a total of 62 persons were involved either as participants or in preparing data sets/evaluating these tasks. Per task, each team was allowed to submit five runs offline and another five online via the BioCreative Meta-Server. From the 52 runs submitted for the ACT, the highest Matthews Correlation Coefficient (MCC) score measured was 0.55 at an accuracy of 89% and the best AUC iP/R was 68%. Most ACT teams explored machine learning methods, some of them also used lexical resources like MeSH terms, PSI-MI concepts or particular lists of verbs and nouns, some integrated NER approaches. For the IMT, a total of 42 runs were evaluated by comparing systems against manually generated annotations done by curators from the BioGRID and MINT databases. The highest AUC iP/R achieved by any run was 53%, the best MCC score 0.55. In case of competitive systems with an acceptable recall (above 35%) the macro-averaged precision ranged between 50% and 80%, with a maximum F-Score of 55%.ConclusionsThe results of the ACT task of BioCreative III indicate that classification of large unbalanced article collections reflecting the real class imbalance is still challenging. Nevertheless, text-mining tools that report ranked lists of relevant articles for manual selection can potentially reduce the time needed to identify half of the relevant articles to less than 1/4 of the time when compared to unranked results. Detecting associations between full text articles and interaction detection method PSI-MI terms (IMT) is more difficult than might be anticipated. This is due to the variability of method term mentions, errors resulting from pre-processing of articles provided as PDF files, and the heterogeneity and different granularity of method term concepts encountered in the ontology. However, combining the sophisticated techniques developed by the participants with supporting evidence strings derived from the articles for human interpretation could result in practical modules for biological annotation workflows.

The CHEMDNER corpus of chemicals and drugs and its annotation principles

The automatic extraction of chemical information from text requires the recognition of chemical entity mentions as one of its key steps. When developing supervised named entity recognition (NER) systems, the availability of a large, manually annotated text corpus is desirable. Furthermore, large corpora permit the robust evaluation and comparison of different approaches that detect chemicals in documents. We present the CHEMDNER corpus, a collection of 10,000 PubMed abstracts that contain a total of 84,355 chemical entity mentions labeled manually by expert chemistry literature curators, following annotation guidelines specifically defined for this task. The abstracts of the CHEMDNER corpus were selected to be representative for all major chemical disciplines. Each of the chemical entity mentions was manually labeled according to its structure-associated chemical entity mention (SACEM) class: abbreviation, family, formula, identifier, multiple, systematic and trivial. The difficulty and consistency of tagging chemicals in text was measured using an agreement study between annotators, obtaining a percentage agreement of 91. For a subset of the CHEMDNER corpus (the test set of 3,000 abstracts) we provide not only the Gold Standard manual annotations, but also mentions automatically detected by the 26 teams that participated in the BioCreative IV CHEMDNER chemical mention recognition task. In addition, we release the CHEMDNER silver standard corpus of automatically extracted mentions from 17,000 randomly selected PubMed abstracts. A version of the CHEMDNER corpus in the BioC format has been generated as well. We propose a standard for required minimum information about entity annotations for the construction of domain specific corpora on chemical and drug entities. The CHEMDNER corpus and annotation guidelines are available at: http://www.biocreative.org/resources/biocreative-iv/chemdner-corpus/

An Overview of BioCreative II.5

We present the results of the BioCreative II.5 evaluation in association with the FEBS Letters experiment, where authors created Structured Digital Abstracts to capture information about protein-protein interactions. The BioCreative II.5 challenge evaluated automatic annotations from 15 text mining teams based on a gold standard created by reconciling annotations from curators, authors, and automated systems. The tasks were to rank articles for curation based on curatable protein-protein interactions; to identify the interacting proteins (using UniProt identifiers) in the positive articles (61); and to identify interacting protein pairs. There were 595 full-text articles in the evaluation test set, including those both with and without curatable protein interactions. The principal evaluation metrics were the interpolated area under the precision/recall curve (AUC iP/R), and (balanced) F-measure. For article classification, the best AUC iP/R was 0.70; for interacting proteins, the best system achieved good macroaveraged recall (0.73) and interpolated area under the precision/recall curve (0.58), after filtering incorrect species and mapping homonymous orthologs; for interacting protein pairs, the top (filtered, mapped) recall was 0.42 and AUC iP/R was 0.29. Ensemble systems improved performance for the interacting protein task.

BioC: a minimalist approach to interoperability for biomedical text processing

A vast amount of scientific information is encoded in natural language text, and the quantity of such text has become so great that it is no longer economically feasible to have a human as the first step in the search process. Natural language processing and text mining tools have become essential to facilitate the search for and extraction of information from text. This has led to vigorous research efforts to create useful tools and to create humanly labeled text corpora, which can be used to improve such tools. To encourage combining these efforts into larger, more powerful and more capable systems, a common interchange format to represent, store and exchange the data in a simple manner between different language processing systems and text mining tools is highly desirable. Here we propose a simple extensible mark-up language format to share text documents and annotations. The proposed annotation approach allows a large number of different annotations to be represented including sentences, tokens, parts of speech, named entities such as genes or diseases and relationships between named entities. In addition, we provide simple code to hold this data, read it from and write it back to extensible mark-up language files and perform some sample processing. We also describe completed as well as ongoing work to apply the approach in several directions. Code and data are available at http://bioc.sourceforge.net/. Database URL: http://bioc.sourceforge.net/

Analysis of Biological Processes and Diseases Using Text Mining Approaches

A number of biomedical text mining systems have been developed to extract biologically relevant information directly from the literature, complementing bioinformatics methods in the analysis of experimentally generated data. We provide a short overview of the general characteristics of natural language data, existing biomedical literature databases, and lexical resources relevant in the context of biomedical text mining. A selected number of practically useful systems are introduced together with the type of user queries supported and the results they generate. The extraction of biological relationships, such as protein-protein interactions as well as metabolic and signaling pathways using information extraction systems, will be discussed through example cases of cancer-relevant proteins. Basic strategies for detecting associations of genes to diseases together with literature mining of mutations, SNPs, and epigenetic information (methylation) are described. We provide an overview of disease-centric and gene-centric literature mining methods for linking genes to phenotypic and genotypic aspects. Moreover, we discuss recent efforts for finding biomarkers through text mining and for gene list analysis and prioritization. Some relevant issues for implementing a customized biomedical text mining system will be pointed out. To demonstrate the usefulness of literature mining for the molecular oncology domain, we implemented two cancer-related applications. The first tool consists of a literature mining system for retrieving human mutations together with supporting articles. Specific gene mutations are linked to a set of predefined cancer types. The second application consists of a text categorization system supporting breast cancer-specific literature search and document-based breast cancer gene ranking. Future trends in text mining emphasize the importance of community efforts such as the BioCreative challenge for the development and integration of multiple systems into a common platform provided by the BioCreative Metaserver.

CHEMDNER: The drugs and chemical names extraction challenge

Natural language processing (NLP) and text mining technologies for the chemical domain (ChemNLP or chemical text mining) are key to improve the access and integration of information from unstructured data such as patents or the scientific literature. Therefore, the BioCreative organizers posed the CHEMDNER (chemical compound and drug name recognition) community challenge, which promoted the development of novel, competitive and accessible chemical text mining systems. This task allowed a comparative assessment of the performance of various methodologies using a carefully prepared collection of manually labeled text prepared by specially trained chemists as Gold Standard data. We evaluated two important aspects: one covered the indexing of documents with chemicals (chemical document indexing - CDI task), and the other was concerned with finding the exact mentions of chemicals in text (chemical entity mention recognition - CEM task). 27 teams (23 academic and 4 commercial, a total of 87 researchers) returned results for the CHEMDNER tasks: 26 teams for CEM and 23 for the CDI task. Top scoring teams obtained an F-score of 87.39% for the CEM task and 88.20% for the CDI task, a very promising result when compared to the agreement between human annotators (91%). The strategies used to detect chemicals included machine learning methods (e.g. conditional random fields) using a variety of features, chemistry and drug lexica, and domain-specific rules. We expect that the tools and resources resulting from this effort will have an impact in future developments of chemical text mining applications and will form the basis to find related chemical information for the detected entities, such as toxicological or pharmacogenomic properties.

Text Mining for Drugs and Chemical Compounds: Methods, Tools and Applications

Providing prior knowledge about biological properties of chemicals, such as kinetic values, protein targets, or toxic effects, can facilitate many aspects of drug development. Chemical information is rapidly accumulating in all sorts of free text documents like patents, industry reports, or scientific articles, which has motivated the development of specifically tailored text mining applications. Despite the potential gains, chemical text mining still faces significant challenges. One of the most salient is the recognition of chemical entities mentioned in text. To help practitioners contribute to this area, a good portion of this review is devoted to this issue, and presents the basic concepts and principles underlying the main strategies. The technical details are introduced and accompanied by relevant bibliographic references. Other tasks discussed are retrieving relevant articles, identifying relationships between chemicals and other entities, or determining the chemical structures of chemicals mentioned in text. This review also introduces a number of published applications that can be used to build pipelines in topics like drug side effects, toxicity, and protein‐disease‐compound network analysis. We conclude the review with an outlook on how we expect the field to evolve, discussing its possibilities and its current limitations.

A text-mining perspective on the requirements for electronically annotated abstracts

We propose that the combination of human expertise and automatic text‐mining systems can be used to create a first generation of electronically annotated information (EAI) that can be added to journal abstracts and that is directly related to the information in the corresponding text. The first experiments have concentrated on the annotation of gene/protein names and those of organisms, as these are the best resolved problems. A second generation of systems could then attempt to address the problems of annotating protein interactions and protein/gene functions, a more difficult task for text‐mining systems. EAI will permit easier categorization of this information, it will help in the evaluation of papers for their curation in databases, and it will be invaluable for maintaining the links between the information in databases and the facts described in text. Additionally, it will contribute to the efforts towards completing database information and creating collections of annotated text that can be used to train new generations of text‐mining systems. The recent introduction of the first meta‐server for the annotation of biological text, with the possibility of collecting annotations from available text‐mining systems, adds credibility to the technical feasibility of this proposal.

How to link ontologies and protein–protein interactions to literature: text-mining approaches and the BioCreative experience

There is an increasing interest in developing ontologies and controlled vocabularies to improve the efficiency and consistency of manual literature curation, to enable more formal biocuration workflow results and ultimately to improve analysis of biological data. Two ontologies that have been successfully used for this purpose are the Gene Ontology (GO) for annotating aspects of gene products and the Molecular Interaction ontology (PSI-MI) used by databases that archive protein–protein interactions. The examination of protein interactions has proven to be extremely promising for the understanding of cellular processes. Manual mapping of information from the biomedical literature to bio-ontology terms is one of the most challenging components in the curation pipeline. It requires that expert curators interpret the natural language descriptions contained in articles and infer their semantic equivalents in the ontology (controlled vocabulary). Since manual curation is a time-consuming process, there is strong motivation to implement text-mining techniques to automatically extract annotations from free text. A range of text mining strategies has been devised to assist in the automated extraction of biological data. These strategies either recognize technical terms used recurrently in the literature and propose them as candidates for inclusion in ontologies, or retrieve passages that serve as evidential support for annotating an ontology term, e.g. from the PSI-MI or GO controlled vocabularies. Here, we provide a general overview of current text-mining methods to automatically extract annotations of GO and PSI-MI ontology terms in the context of the BioCreative (Critical Assessment of Information Extraction Systems in Biology) challenge. Special emphasis is given to protein–protein interaction data and PSI-MI terms referring to interaction detection methods.