Integrating high dimensional bi-directional parsing models for gene mention tagging

Exploiting and assessing multi-source data for supervised biomedical named entity recognition

Motivation

Recognition of biomedical entities from scientific text is a critical component of natural language processing and automated information extraction platforms. Modern named entity recognition approaches rely heavily on supervised machine learning techniques, which are critically dependent on annotated training corpora. These approaches have been shown to perform well when trained and tested on the same source. However, in such scenario, the performance and evaluation of these models may be optimistic, as such models may not necessarily generalize to independent corpora, resulting in potential non-optimal entity recognition for large-scale tagging of widely diverse articles in databases such as PubMed.

Results

Here we aggregated published corpora for the recognition of biomolecular entities (such as genes, RNA, proteins, variants, drugs and metabolites), identified entity class overlap and performed leave-corpus-out cross validation strategy to test the efficiency of existing models. We demonstrate that accuracies of models trained on individual corpora decrease substantially for recognition of the same biomolecular entity classes in independent corpora. This behavior is possibly due to limited generalizability of entity-class-related features captured by individual corpora (model 'overtraining') which we investigated further at the orthographic level, as well as potential annotation standard differences. We show that the combined use of multi-source training corpora results in overall more generalizable models for named entity recognition, while achieving comparable individual performance. By performing learning-curve-based power analysis we further identified that performance is often not limited by the quantity of the annotated data.

Availability and implementation

Compiled primary and secondary sources of the aggregated corpora are available on: https://github.com/dterg/biomedical_corpora/wiki and https://bitbucket.org/iAnalytica/bioner.

Supplementary information

Supplementary data are available at Bioinformatics online.

GRAM-CNN: a deep learning approach with local context for named entity recognition in biomedical text

Motivation

Best performing named entity recognition (NER) methods for biomedical literature are based on hand-crafted features or task-specific rules, which are costly to produce and difficult to generalize to other corpora. End-to-end neural networks achieve state-of-the-art performance without hand-crafted features and task-specific knowledge in non-biomedical NER tasks. However, in the biomedical domain, using the same architecture does not yield competitive performance compared with conventional machine learning models.

Results

We propose a novel end-to-end deep learning approach for biomedical NER tasks that leverages the local contexts based on n-gram character and word embeddings via Convolutional Neural Network (CNN). We call this approach GRAM-CNN. To automatically label a word, this method uses the local information around a word. Therefore, the GRAM-CNN method does not require any specific knowledge or feature engineering and can be theoretically applied to a wide range of existing NER problems. The GRAM-CNN approach was evaluated on three well-known biomedical datasets containing different BioNER entities. It obtained an F1-score of 87.26% on the Biocreative II dataset, 87.26% on the NCBI dataset and 72.57% on the JNLPBA dataset. Those results put GRAM-CNN in the lead of the biological NER methods. To the best of our knowledge, we are the first to apply CNN based structures to BioNER problems.

Availability and implementation

The GRAM-CNN source code, datasets and pre-trained model are available online at: https://github.com/valdersoul/GRAM-CNN.

Contact

andyli@ece.ufl.edu or aconesa@ufl.edu.

Supplementary information

Supplementary data are available at Bioinformatics online.

Extraction of Information Related to Adverse Drug Events from Electronic Health Record Notes: Design of an End-to-End Model Based on Deep Learning

Background

Pharmacovigilance and drug-safety surveillance are crucial for monitoring adverse drug events (ADEs), but the main ADE-reporting systems such as Food and Drug Administration Adverse Event Reporting System face challenges such as underreporting. Therefore, as complementary surveillance, data on ADEs are extracted from electronic health record (EHR) notes via natural language processing (NLP). As NLP develops, many up-to-date machine-learning techniques are introduced in this field, such as deep learning and multi-task learning (MTL). However, only a few studies have focused on employing such techniques to extract ADEs.

Objective

We aimed to design a deep learning model for extracting ADEs and related information such as medications and indications. Since extraction of ADE-related information includes two steps—named entity recognition and relation extraction—our second objective was to improve the deep learning model using multi-task learning between the two steps.

Methods

We employed the dataset from the Medication, Indication and Adverse Drug Events (MADE) 1.0 challenge to train and test our models. This dataset consists of 1089 EHR notes of cancer patients and includes 9 entity types such as Medication, Indication, and ADE and 7 types of relations between these entities. To extract information from the dataset, we proposed a deep-learning model that uses a bidirectional long short-term memory (BiLSTM) conditional random field network to recognize entities and a BiLSTM-Attention network to extract relations. To further improve the deep-learning model, we employed three typical MTL methods, namely, hard parameter sharing, parameter regularization, and task relation learning, to build three MTL models, called HardMTL, RegMTL, and LearnMTL, respectively.

Results

Since extraction of ADE-related information is a two-step task, the result of the second step (ie, relation extraction) was used to compare all models. We used microaveraged precision, recall, and F1 as evaluation metrics. Our deep learning model achieved state-of-the-art results (F1=65.9%), which is significantly higher than that (F1=61.7%) of the best system in the MADE1.0 challenge. HardMTL further improved the F1 by 0.8%, boosting the F1 to 66.7%, whereas RegMTL and LearnMTL failed to boost the performance.

Conclusions

Deep learning models can significantly improve the performance of ADE-related information extraction. MTL may be effective for named entity recognition and relation extraction, but it depends on the methods, data, and other factors. Our results can facilitate research on ADE detection, NLP, and machine learning.

Long short-term memory RNN for biomedical named entity recognition

Background

Biomedical named entity recognition(BNER) is a crucial initial step of information extraction in biomedical domain. The task is typically modeled as a sequence labeling problem. Various machine learning algorithms, such as Conditional Random Fields (CRFs), have been successfully used for this task. However, these state-of-the-art BNER systems largely depend on hand-crafted features.

Results

We present a recurrent neural network (RNN) framework based on word embeddings and character representation. On top of the neural network architecture, we use a CRF layer to jointly decode labels for the whole sentence. In our approach, contextual information from both directions and long-range dependencies in the sequence, which is useful for this task, can be well modeled by bidirectional variation and long short-term memory (LSTM) unit, respectively. Although our models use word embeddings and character embeddings as the only features, the bidirectional LSTM-RNN (BLSTM-RNN) model achieves state-of-the-art performance — 86.55% F1 on BioCreative II gene mention (GM) corpus and 73.79% F1 on JNLPBA 2004 corpus.

Conclusions

Our neural network architecture can be successfully used for BNER without any manual feature engineering. Experimental results show that domain-specific pre-trained word embeddings and character-level representation can improve the performance of the LSTM-RNN models. On the GM corpus, we achieve comparable performance compared with other systems using complex hand-crafted features. Considering the JNLPBA corpus, our model achieves the best results, outperforming the previously top performing systems. The source code of our method is freely available under GPL at https://github.com/lvchen1989/BNER.

Semantic annotation in biomedicine: the current landscape

The abundance and unstructured nature of biomedical texts, be it clinical or research content, impose significant challenges for the effective and efficient use of information and knowledge stored in such texts. Annotation of biomedical documents with machine intelligible semantics facilitates advanced, semantics-based text management, curation, indexing, and search. This paper focuses on annotation of biomedical entity mentions with concepts from relevant biomedical knowledge bases such as UMLS. As a result, the meaning of those mentions is unambiguously and explicitly defined, and thus made readily available for automated processing. This process is widely known as semantic annotation, and the tools that perform it are known as semantic annotators.Over the last dozen years, the biomedical research community has invested significant efforts in the development of biomedical semantic annotation technology. Aiming to establish grounds for further developments in this area, we review a selected set of state of the art biomedical semantic annotators, focusing particularly on general purpose annotators, that is, semantic annotation tools that can be customized to work with texts from any area of biomedicine. We also examine potential directions for further improvements of today's annotators which could make them even more capable of meeting the needs of real-world applications. To motivate and encourage further developments in this area, along the suggested and/or related directions, we review existing and potential practical applications and benefits of semantic annotators.

RysannMD: A biomedical semantic annotator balancing speed and accuracy

Recently, both researchers and practitioners have explored the possibility of semantically annotating large and continuously evolving collections of biomedical texts such as research papers, medical reports, and physician notes in order to enable their efficient and effective management and use in clinical practice or research laboratories. Such annotations can be automatically generated by biomedical semantic annotators - tools that are specifically designed for detecting and disambiguating biomedical concepts mentioned in text. The biomedical community has already presented several solid automated semantic annotators. However, the existing tools are either strong in their disambiguation capacity, i.e., the ability to identify the correct biomedical concept for a given piece of text among several candidate concepts, or they excel in their processing time, i.e., work very efficiently, but none of the semantic annotation tools reported in the literature has both of these qualities. In this paper, we present RysannMD (Ryerson Semantic Annotator for Medical Domain), a biomedical semantic annotation tool that strikes a balance between processing time and performance while disambiguating biomedical terms. In other words, RysannMD provides reasonable disambiguation performance when choosing the right sense for a biomedical term in a given context, and does that in a reasonable time. To examine how RysannMD stands with respect to the state of the art biomedical semantic annotators, we have conducted a series of experiments using standard benchmarking corpora, including both gold and silver standards, and four modern biomedical semantic annotators, namely cTAKES, MetaMap, NOBLE Coder, and Neji. The annotators were compared with respect to the quality of the produced annotations measured against gold and silver standards using precision, recall, and F₁ measure and speed, i.e., processing time. In the experiments, RysannMD achieved the best median F₁ measure across the benchmarking corpora, independent of the standard used (silver/gold), biomedical subdomain, and document size. In terms of the annotation speed, RysannMD scored the second best median processing time across all the experiments. The obtained results indicate that RysannMD offers the best performance among the examined semantic annotators when both quality of annotation and speed are considered simultaneously.

BCC-NER: bidirectional, contextual clues named entity tagger for gene/protein mention recognition

Tagging biomedical entities such as gene, protein, cell, and cell-line is the first step and an important pre-requisite in biomedical literature mining. In this paper, we describe our hybrid named entity tagging approach namely BCC-NER (bidirectional, contextual clues named entity tagger for gene/protein mention recognition). BCC-NER is deployed with three modules. The first module is for text processing which includes basic NLP pre-processing, feature extraction, and feature selection. The second module is for training and model building with bidirectional conditional random fields (CRF) to parse the text in both directions (forward and backward) and integrate the backward and forward trained models using margin-infused relaxed algorithm (MIRA). The third and final module is for post-processing to achieve a better performance, which includes surrounding text features, parenthesis mismatching, and two-tier abbreviation algorithm. The evaluation results on BioCreative II GM test corpus of BCC-NER achieve a precision of 89.95, recall of 84.15 and overall F-score of 86.95, which is higher than the other currently available open source taggers.

Automated Neuroanatomical Relation Extraction: A Linguistically Motivated Approach with a PVT Connectivity Graph Case Study

Identifying the relations among different regions of the brain is vital for a better understanding of how the brain functions. While a large number of studies have investigated the neuroanatomical and neurochemical connections among brain structures, their specific findings are found in publications scattered over a large number of years and different types of publications. Text mining techniques have provided the means to extract specific types of information from a large number of publications with the aim of presenting a larger, if not necessarily an exhaustive picture. By using natural language processing techniques, the present paper aims to identify connectivity relations among brain regions in general and relations relevant to the paraventricular nucleus of the thalamus (PVT) in particular. We introduce a linguistically motivated approach based on patterns defined over the constituency and dependency parse trees of sentences. Besides the presence of a relation between a pair of brain regions, the proposed method also identifies the directionality of the relation, which enables the creation and analysis of a directional brain region connectivity graph. The approach is evaluated over the manually annotated data sets of the WhiteText Project. In addition, as a case study, the method is applied to extract and analyze the connectivity graph of PVT, which is an important brain region that is considered to influence many functions ranging from arousal, motivation, and drug-seeking behavior to attention. The results of the PVT connectivity graph show that PVT may be a new target of research in mood assessment.

Finding Related Publications: Extending the Set of Terms Used to Assess Article Similarity

Recommendation of related articles is an important feature of the PubMed. The PubMed Related Citations (PRC) algorithm is the engine that enables this feature, and it leverages information on 22 million citations. We analyzed the performance of the PRC algorithm on 4584 annotated articles from the 2005 Text REtrieval Conference (TREC) Genomics Track data. Our analysis indicated that the PRC highest weighted term was not always consistent with the critical term that was most directly related to the topic of the article. We implemented term expansion and found that it was a promising and easy-to-implement approach to improve the performance of the PRC algorithm for the TREC 2005 Genomics data and for the TREC 2014 Clinical Decision Support Track data. For term expansion, we trained a Skip-gram model using the Word2Vec package. This extended PRC algorithm resulted in higher average precision for a large subset of articles. A combination of both algorithms may lead to improved performance in related article recommendations.

AuDis: an automatic CRF-enhanced disease normalization in biomedical text

Diseases play central roles in many areas of biomedical research and healthcare. Consequently, aggregating the disease knowledge and treatment research reports becomes an extremely critical issue, especially in rapid-growth knowledge bases (e.g. PubMed). We therefore developed a system, AuDis, for disease mention recognition and normalization in biomedical texts. Our system utilizes an order two conditional random fields model. To optimize the results, we customize several post-processing steps, including abbreviation resolution, consistency improvement and stopwords filtering. As the official evaluation on the CDR task in BioCreative V, AuDis obtained the best performance (86.46% of F-score) among 40 runs (16 unique teams) on disease normalization of the DNER sub task. These results suggest that AuDis is a high-performance recognition system for disease recognition and normalization from biomedical literature.Database URL: http://ikmlab.csie.ncku.edu.tw/CDR2015/AuDis.html.

Mining chemical patents with an ensemble of open systems

The significant amount of medicinal chemistry information contained in patents makes them an attractive target for text mining. In this manuscript, we describe systems for named entity recognition (NER) of chemicals and genes/proteins in patents, using the CEMP (for chemicals) and GPRO (for genes/proteins) corpora provided by the CHEMDNER task at BioCreative V. Our chemical NER system is an ensemble of five open systems, including both versions of tmChem, our previous work on chemical NER. Their output is combined using a machine learning classification approach. Our chemical NER system obtained 0.8752 precision and 0.9129 recall, for 0.8937 f-score on the CEMP task. Our gene/protein NER system is an extension of our previous work for gene and protein NER, GNormPlus. This system obtained a performance of 0.8143 precision and 0.8141 recall, for 0.8137 f-score on the GPRO task. Both systems achieved the highest performance in their respective tasks at BioCreative V. We conclude that an ensemble of independently-created open systems is sufficiently diverse to significantly improve performance over any individual system, even when they use a similar approach.

Database URL: http://www.ncbi.nlm.nih.gov/CBBresearch/Lu/Demo/tmTools/.

NOBLE - Flexible concept recognition for large-scale biomedical natural language processing

Background

Natural language processing (NLP) applications are increasingly important in biomedical data analysis, knowledge engineering, and decision support. Concept recognition is an important component task for NLP pipelines, and can be either general-purpose or domain-specific. We describe a novel, flexible, and general-purpose concept recognition component for NLP pipelines, and compare its speed and accuracy against five commonly used alternatives on both a biological and clinical corpus.

NOBLE Coder implements a general algorithm for matching terms to concepts from an arbitrary vocabulary set. The system’s matching options can be configured individually or in combination to yield specific system behavior for a variety of NLP tasks. The software is open source, freely available, and easily integrated into UIMA or GATE. We benchmarked speed and accuracy of the system against the CRAFT and ShARe corpora as reference standards and compared it to MMTx, MGrep, Concept Mapper, cTAKES Dictionary Lookup Annotator, and cTAKES Fast Dictionary Lookup Annotator.

Results

We describe key advantages of the NOBLE Coder system and associated tools, including its greedy algorithm, configurable matching strategies, and multiple terminology input formats. These features provide unique functionality when compared with existing alternatives, including state-of-the-art systems. On two benchmarking tasks, NOBLE’s performance exceeded commonly used alternatives, performing almost as well as the most advanced systems. Error analysis revealed differences in error profiles among systems.

Conclusion

NOBLE Coder is comparable to other widely used concept recognition systems in terms of accuracy and speed. Advantages of NOBLE Coder include its interactive terminology builder tool, ease of configuration, and adaptability to various domains and tasks. NOBLE provides a term-to-concept matching system suitable for general concept recognition in biomedical NLP pipelines.

Literature Mining and Ontology based Analysis of Host-Brucella Gene-Gene Interaction Network

Brucella is an intracellular bacterium that causes chronic brucellosis in humans and various mammals. The identification of host-Brucella interaction is crucial to understand host immunity against Brucella infection and Brucella pathogenesis against host immune responses. Most of the information about the inter-species interactions between host and Brucella genes is only available in the text of the scientific publications. Many text-mining systems for extracting gene and protein interactions have been proposed. However, only a few of them have been designed by considering the peculiarities of host–pathogen interactions. In this paper, we used a text mining approach for extracting host-Brucella gene–gene interactions from the abstracts of articles in PubMed. The gene–gene interactions here represent the interactions between genes and/or gene products (e.g., proteins). The SciMiner tool, originally designed for detecting mammalian gene/protein names in text, was extended to identify host and Brucella gene/protein names in the abstracts. Next, sentence-level and abstract-level co-occurrence based approaches, as well as sentence-level machine learning based methods, originally designed for extracting intra-species gene interactions, were utilized to extract the interactions among the identified host and Brucella genes. The extracted interactions were manually evaluated. A total of 46 host-Brucella gene interactions were identified and represented as an interaction network. Twenty four of these interactions were identified from sentence-level processing. Twenty two additional interactions were identified when abstract-level processing was performed. The Interaction Network Ontology (INO) was used to represent the identified interaction types at a hierarchical ontology structure. Ontological modeling of specific gene–gene interactions demonstrates that host–pathogen gene–gene interactions occur at experimental conditions which can be ontologically represented. Our results show that the introduced literature mining and ontology-based modeling approach are effective in retrieving and analyzing host–pathogen gene–gene interaction networks.

An Overview of Biomolecular Event Extraction from Scientific Documents

This paper presents a review of state-of-the-art approaches to automatic extraction of biomolecular events from scientific texts. Events involving biomolecules such as genes, transcription factors, or enzymes, for example, have a central role in biological processes and functions and provide valuable information for describing physiological and pathogenesis mechanisms. Event extraction from biomedical literature has a broad range of applications, including support for information retrieval, knowledge summarization, and information extraction and discovery. However, automatic event extraction is a challenging task due to the ambiguity and diversity of natural language and higher-level linguistic phenomena, such as speculations and negations, which occur in biological texts and can lead to misunderstanding or incorrect interpretation. Many strategies have been proposed in the last decade, originating from different research areas such as natural language processing, machine learning, and statistics. This review summarizes the most representative approaches in biomolecular event extraction and presents an analysis of the current state of the art and of commonly used methods, features, and tools. Finally, current research trends and future perspectives are also discussed.

GNormPlus: An Integrative Approach for Tagging Genes, Gene Families, and Protein Domains

The automatic recognition of gene names and their associated database identifiers from biomedical text has been widely studied in recent years, as these tasks play an important role in many downstream text-mining applications. Despite significant previous research, only a small number of tools are publicly available and these tools are typically restricted to detecting only mention level gene names or only document level gene identifiers. In this work, we report GNormPlus: an end-to-end and open source system that handles both gene mention and identifier detection. We created a new corpus of 694 PubMed articles to support our development of GNormPlus, containing manual annotations for not only gene names and their identifiers, but also closely related concepts useful for gene name disambiguation, such as gene families and protein domains. GNormPlus integrates several advanced text-mining techniques, including SimConcept for resolving composite gene names. As a result, GNormPlus compares favorably to other state-of-the-art methods when evaluated on two widely used public benchmarking datasets, achieving 86.7% F1-score on the BioCreative II Gene Normalization task dataset and 50.1% F1-score on the BioCreative III Gene Normalization task dataset. The GNormPlus source code and its annotated corpus are freely available, and the results of applying GNormPlus to the entire PubMed are freely accessible through our web-based tool PubTator.

PKDE4J: Entity and relation extraction for public knowledge discovery

Due to an enormous number of scientific publications that cannot be handled manually, there is a rising interest in text-mining techniques for automated information extraction, especially in the biomedical field. Such techniques provide effective means of information search, knowledge discovery, and hypothesis generation. Most previous studies have primarily focused on the design and performance improvement of either named entity recognition or relation extraction. In this paper, we present PKDE4J, a comprehensive text-mining system that integrates dictionary-based entity extraction and rule-based relation extraction in a highly flexible and extensible framework. Starting with the Stanford CoreNLP, we developed the system to cope with multiple types of entities and relations. The system also has fairly good performance in terms of accuracy as well as the ability to configure text-processing components. We demonstrate its competitive performance by evaluating it on many corpora and found that it surpasses existing systems with average F-measures of 85% for entity extraction and 81% for relation extraction.

pGenN, a gene normalization tool for plant genes and proteins in scientific literature

BACKGROUND

Automatically detecting gene/protein names in the literature and connecting them to databases records, also known as gene normalization, provides a means to structure the information buried in free-text literature. Gene normalization is critical for improving the coverage of annotation in the databases, and is an essential component of many text mining systems and database curation pipelines.

METHODS

In this manuscript, we describe a gene normalization system specifically tailored for plant species, called pGenN (pivot-based Gene Normalization). The system consists of three steps: dictionary-based gene mention detection, species assignment, and intra species normalization. We have developed new heuristics to improve each of these phases.

RESULTS

We evaluated the performance of pGenN on an in-house expertly annotated corpus consisting of 104 plant relevant abstracts. Our system achieved an F-value of 88.9% (Precision 90.9% and Recall 87.2%) on this corpus, outperforming state-of-art systems presented in BioCreative III. We have processed over 440,000 plant-related Medline abstracts using pGenN. The gene normalization results are stored in a local database for direct query from the pGenN web interface (proteininformationresource.org/pgenn/). The annotated literature corpus is also publicly available through the PIR text mining portal (proteininformationresource.org/iprolink/).

Curatable Named-Entity Recognition Using Semantic Relations

Named-entity recognition (NER) plays an important role in the development of biomedical databases. However, the existing NER tools produce multifarious named-entities which may result in both curatable and non-curatable markers. To facilitate biocuration with a straightforward approach, classifying curatable named-entities is helpful with regard to accelerating the biocuration workflow. Co-occurrence Interaction Nexus with Named-entity Recognition (CoINNER) is a web-based tool that allows users to identify genes, chemicals, diseases, and action term mentions in the Comparative Toxicogenomic Database (CTD). To further discover interactions, CoINNER uses multiple advanced algorithms to recognize the mentions in the BioCreative IV CTD Track. CoINNER is developed based on a prototype system that annotated gene, chemical, and disease mentions in PubMed abstracts at BioCreative 2012 Track I (literature triage). We extended our previous system in developing CoINNER. The pre-tagging results of CoINNER were developed based on the state-of-the-art named entity recognition tools in BioCreative III. Next, a method based on conditional random fields (CRFs) is proposed to predict chemical and disease mentions in the articles. Finally, action term mentions were collected by latent Dirichlet allocation (LDA). At the BioCreative IV CTD Track, the best F-measures reached for gene/protein, chemical/drug and disease NER were 54 percent while CoINNER achieved a 61.5 percent F-measure. System URL: http://ikmbio.csie.ncku.edu.tw/coinner/ introduction.htm.

A review on computational systems biology of pathogen-host interactions

Pathogens manipulate the cellular mechanisms of host organisms via pathogen-host interactions (PHIs) in order to take advantage of the capabilities of host cells, leading to infections. The crucial role of these interspecies molecular interactions in initiating and sustaining infections necessitates a thorough understanding of the corresponding mechanisms. Unlike the traditional approach of considering the host or pathogen separately, a systems-level approach, considering the PHI system as a whole is indispensable to elucidate the mechanisms of infection. Following the technological advances in the post-genomic era, PHI data have been produced in large-scale within the last decade. Systems biology-based methods for the inference and analysis of PHI regulatory, metabolic, and protein-protein networks to shed light on infection mechanisms are gaining increasing demand thanks to the availability of omics data. The knowledge derived from the PHIs may largely contribute to the identification of new and more efficient therapeutics to prevent or cure infections. There are recent efforts for the detailed documentation of these experimentally verified PHI data through Web-based databases. Despite these advances in data archiving, there are still large amounts of PHI data in the biomedical literature yet to be discovered, and novel text mining methods are in development to unearth such hidden data. Here, we review a collection of recent studies on computational systems biology of PHIs with a special focus on the methods for the inference and analysis of PHI networks, covering also the Web-based databases and text-mining efforts to unravel the data hidden in the literature.

Enhancing medical named entity recognition with an extended segment representation technique

OBJECTIVE

The objective of this paper is to formulate an extended segment representation (SR) technique to enhance named entity recognition (NER) in medical applications.

METHODS

An extension to the IOBES (Inside/Outside/Begin/End/Single) SR technique is formulated. In the proposed extension, a new class is assigned to words that do not belong to a named entity (NE) in one context but appear as an NE in other contexts. Ambiguity in such cases can negatively affect the results of classification-based NER techniques. Assigning a separate class to words that can potentially cause ambiguity in NER allows a classifier to detect NEs more accurately; therefore increasing classification accuracy.

RESULTS

The proposed SR technique is evaluated using the i2b2 2010 medical challenge data set with eight different classifiers. Each classifier is trained separately to extract three different medical NEs, namely treatment, problem, and test. From the three experimental results, the extended SR technique is able to improve the average F1-measure results pertaining to seven out of eight classifiers. The kNN classifier shows an average reduction of 0.18% across three experiments, while the C4.5 classifier records an average improvement of 9.33%.

Enhancing of chemical compound and drug name recognition using representative tag scheme and fine-grained tokenization

Background

The functions of chemical compounds and drugs that affect biological processes and their particular effect on the onset and treatment of diseases have attracted increasing interest with the advancement of research in the life sciences. To extract knowledge from the extensive literatures on such compounds and drugs, the organizers of BioCreative IV administered the CHEMical Compound and Drug Named Entity Recognition (CHEMDNER) task to establish a standard dataset for evaluating state-of-the-art chemical entity recognition methods.

Methods

This study introduces the approach of our CHEMDNER system. Instead of emphasizing the development of novel feature sets for machine learning, this study investigates the effect of various tag schemes on the recognition of the names of chemicals and drugs by using conditional random fields. Experiments were conducted using combinations of different tokenization strategies and tag schemes to investigate the effects of tag set selection and tokenization method on the CHEMDNER task.

Results

This study presents the performance of CHEMDNER of three more representative tag schemes-IOBE, IOBES, and IOB₁₂E-when applied to a widely utilized IOB tag set and combined with the coarse-/fine-grained tokenization methods. The experimental results thus reveal that the fine-grained tokenization strategy performance best in terms of precision, recall and F-scores when the IOBES tag set was utilized. The IOBES model with fine-grained tokenization yielded the best-F-scores in the six chemical entity categories other than the "Multiple" entity category. Nonetheless, no significant improvement was observed when a more representative tag schemes was used with the coarse or fine-grained tokenization rules. The best F-scores that were achieved using the developed system on the test dataset of the CHEMDNER task were 0.833 and 0.815 for the chemical documents indexing and the chemical entity mention recognition tasks, respectively.

Conclusions

The results herein highlight the importance of tag set selection and the use of different tokenization strategies. Fine-grained tokenization combined with the tag set IOBES most effectively recognizes chemical and drug names. To the best of the authors' knowledge, this investigation is the first comprehensive investigation use of various tag set schemes combined with different tokenization strategies for the recognition of chemical entities.

tmChem: a high performance approach for chemical named entity recognition and normalization

Chemical compounds and drugs are an important class of entities in biomedical research with great potential in a wide range of applications, including clinical medicine. Locating chemical named entities in the literature is a useful step in chemical text mining pipelines for identifying the chemical mentions, their properties, and their relationships as discussed in the literature.

We introduce the tmChem system, a chemical named entity recognizer created by combining two independent machine learning models in an ensemble. We use the corpus released as part of the recent CHEMDNER task to develop and evaluate tmChem, achieving a micro-averaged f-measure of 0.8739 on the CEM subtask (mention-level evaluation) and 0.8745 f-measure on the CDI subtask (abstract-level evaluation). We also report a high-recall combination (0.9212 for CEM and 0.9224 for CDI). tmChem achieved the highest f-measure reported in the CHEMDNER task for the CEM subtask, and the high recall variant achieved the highest recall on both the CEM and CDI tasks.

We report that tmChem is a state-of-the-art tool for chemical named entity recognition and that performance for chemical named entity recognition has now tied (or exceeded) the performance previously reported for genes and diseases. Future research should focus on tighter integration between the named entity recognition and normalization steps for improved performance.

The source code and a trained model for both models of tmChem is available at: http://www.ncbi.nlm.nih.gov/CBBresearch/Lu/Demo/tmChem. The results of running tmChem (Model 2) on PubMed are available in PubTator: http://www.ncbi.nlm.nih.gov/CBBresearch/Lu/Demo/PubTator

Incorporating domain knowledge in chemical and biomedical named entity recognition with word representations

Background

Chemical and biomedical Named Entity Recognition (NER) is an essential prerequisite task before effective text mining can begin for biochemical-text data. Exploiting unlabeled text data to leverage system performance has been an active and challenging research topic in text mining due to the recent growth in the amount of biomedical literature.

We present a semi-supervised learning method that efficiently exploits unlabeled data in order to incorporate domain knowledge into a named entity recognition model and to leverage system performance. The proposed method includes Natural Language Processing (NLP) tasks for text preprocessing, learning word representation features from a large amount of text data for feature extraction, and conditional random fields for token classification. Other than the free text in the domain, the proposed method does not rely on any lexicon nor any dictionary in order to keep the system applicable to other NER tasks in bio-text data.

Results

We extended BANNER, a biomedical NER system, with the proposed method. This yields an integrated system that can be applied to chemical and drug NER or biomedical NER. We call our branch of the BANNER system BANNER-CHEMDNER, which is scalable over millions of documents, processing about 530 documents per minute, is configurable via XML, and can be plugged into other systems by using the BANNER Unstructured Information Management Architecture (UIMA) interface.

BANNER-CHEMDNER achieved an 85.68% and an 86.47% F-measure on the testing sets of CHEMDNER Chemical Entity Mention (CEM) and Chemical Document Indexing (CDI) subtasks, respectively, and achieved an 87.04% F-measure on the official testing set of the BioCreative II gene mention task, showing remarkable performance in both chemical and biomedical NER. BANNER-CHEMDNER system is available at: https://bitbucket.org/tsendeemts/banner-chemdner.

Hybrid curation of gene-mutation relations combining automated extraction and crowdsourcing

Background: This article describes capture of biological information using a hybrid approach that combines natural language processing to extract biological entities and crowdsourcing with annotators recruited via Amazon Mechanical Turk to judge correctness of candidate biological relations. These techniques were applied to extract gene– mutation relations from biomedical abstracts with the goal of supporting production scale capture of gene–mutation–disease findings as an open source resource for personalized medicine. Results: The hybrid system could be configured to provide good performance for gene–mutation extraction (precision ∼82%; recall ∼70% against an expert-generated gold standard) at a cost of $0.76 per abstract. This demonstrates that crowd labor platforms such as Amazon Mechanical Turk can be used to recruit quality annotators, even in an application requiring subject matter expertise; aggregated Turker judgments for gene–mutation relations exceeded 90% accuracy. Over half of the precision errors were due to mismatches against the gold standard hidden from annotator view (e.g. incorrect EntrezGene identifier or incorrect mutation position extracted), or incomplete task instructions (e.g. the need to exclude nonhuman mutations). Conclusions: The hybrid curation model provides a readily scalable cost-effective approach to curation, particularly if coupled with expert human review to filter precision errors. We plan to generalize the framework and make it available as open source software.

Database URL:http://www.mitre.org/publications/technical-papers/hybrid-curation-of-gene-mutation-relations-combining-automated

Biomedical relation extraction: from binary to complex

Biomedical relation extraction aims to uncover high-quality relations from life science literature with high accuracy and efficiency. Early biomedical relation extraction tasks focused on capturing binary relations, such as protein-protein interactions, which are crucial for virtually every process in a living cell. Information about these interactions provides the foundations for new therapeutic approaches. In recent years, more interests have been shifted to the extraction of complex relations such as biomolecular events. While complex relations go beyond binary relations and involve more than two arguments, they might also take another relation as an argument. In the paper, we conduct a thorough survey on the research in biomedical relation extraction. We first present a general framework for biomedical relation extraction and then discuss the approaches proposed for binary and complex relation extraction with focus on the latter since it is a much more difficult task compared to binary relation extraction. Finally, we discuss challenges that we are facing with complex relation extraction and outline possible solutions and future directions.

A literature search tool for intelligent extraction of disease-associated genes

OBJECTIVE

To extract disorder-associated genes from the scientific literature in PubMed with greater sensitivity for literature-based support than existing methods.

METHODS

We developed a PubMed query to retrieve disorder-related, original research articles. Then we applied a rule-based text-mining algorithm with keyword matching to extract target disorders, genes with significant results, and the type of study described by the article.

RESULTS

We compared our resulting candidate disorder genes and supporting references with existing databases. We demonstrated that our candidate gene set covers nearly all genes in manually curated databases, and that the references supporting the disorder-gene link are more extensive and accurate than other general purpose gene-to-disorder association databases.

CONCLUSIONS

We implemented a novel publication search tool to find target articles, specifically focused on links between disorders and genotypes. Through comparison against gold-standard manually updated gene-disorder databases and comparison with automated databases of similar functionality we show that our tool can search through the entirety of PubMed to extract the main gene findings for human diseases rapidly and accurately.

tagtog: interactive and text-mining-assisted annotation of gene mentions in PLOS full-text articles

The breadth and depth of biomedical literature are increasing year upon year. To keep abreast of these increases, FlyBase, a database for Drosophila genomic and genetic information, is constantly exploring new ways to mine the published literature to increase the efficiency and accuracy of manual curation and to automate some aspects, such as triaging and entity extraction. Toward this end, we present the 'tagtog' system, a web-based annotation framework that can be used to mark up biological entities (such as genes) and concepts (such as Gene Ontology terms) in full-text articles. tagtog leverages manual user annotation in combination with automatic machine-learned annotation to provide accurate identification of gene symbols and gene names. As part of the BioCreative IV Interactive Annotation Task, FlyBase has used tagtog to identify and extract mentions of Drosophila melanogaster gene symbols and names in full-text biomedical articles from the PLOS stable of journals. We show here the results of three experiments with different sized corpora and assess gene recognition performance and curation speed. We conclude that tagtog-named entity recognition improves with a larger corpus and that tagtog-assisted curation is quicker than manual curation. DATABASE URL: www.tagtog.net, www.flybase.org.

CoIN: a network analysis for document triage

In recent years, there was a rapid increase in the number of medical articles. The number of articles in PubMed has increased exponentially. Thus, the workload for biocurators has also increased exponentially. Under these circumstances, a system that can automatically determine in advance which article has a higher priority for curation can effectively reduce the workload of biocurators. Determining how to effectively find the articles required by biocurators has become an important task. In the triage task of BioCreative 2012, we proposed the Co-occurrence Interaction Nexus (CoIN) for learning and exploring relations in articles. We constructed a co-occurrence analysis system, which is applicable to PubMed articles and suitable for gene, chemical and disease queries. CoIN uses co-occurrence features and their network centralities to assess the influence of curatable articles from the Comparative Toxicogenomics Database. The experimental results show that our network-based approach combined with co-occurrence features can effectively classify curatable and non-curatable articles. CoIN also allows biocurators to survey the ranking lists for specific queries without reviewing meaningless information. At BioCreative 2012, CoIN achieved a 0.778 mean average precision in the triage task, thus finishing in second place out of all participants. Database URL: http://ikmbio.csie.ncku.edu.tw/coin/home.php.

Large-scale event extraction from literature with multi-level gene normalization

Text mining for the life sciences aims to aid database curation, knowledge summarization and information retrieval through the automated processing of biomedical texts. To provide comprehensive coverage and enable full integration with existing biomolecular database records, it is crucial that text mining tools scale up to millions of articles and that their analyses can be unambiguously linked to information recorded in resources such as UniProt, KEGG, BioGRID and NCBI databases. In this study, we investigate how fully automated text mining of complex biomolecular events can be augmented with a normalization strategy that identifies biological concepts in text, mapping them to identifiers at varying levels of granularity, ranging from canonicalized symbols to unique gene and proteins and broad gene families. To this end, we have combined two state-of-the-art text mining components, previously evaluated on two community-wide challenges, and have extended and improved upon these methods by exploiting their complementary nature. Using these systems, we perform normalization and event extraction to create a large-scale resource that is publicly available, unique in semantic scope, and covers all 21.9 million PubMed abstracts and 460 thousand PubMed Central open access full-text articles. This dataset contains 40 million biomolecular events involving 76 million gene/protein mentions, linked to 122 thousand distinct genes from 5032 species across the full taxonomic tree. Detailed evaluations and analyses reveal promising results for application of this data in database and pathway curation efforts. The main software components used in this study are released under an open-source license. Further, the resulting dataset is freely accessible through a novel API, providing programmatic and customized access (http://www.evexdb.org/api/v001/). Finally, to allow for large-scale bioinformatic analyses, the entire resource is available for bulk download from http://evexdb.org/download/, under the Creative Commons – Attribution – Share Alike (CC BY-SA) license.

tmVar: a text mining approach for extracting sequence variants in biomedical literature

MOTIVATION

Text-mining mutation information from the literature becomes a critical part of the bioinformatics approach for the analysis and interpretation of sequence variations in complex diseases in the post-genomic era. It has also been used for assisting the creation of disease-related mutation databases. Most of existing approaches are rule-based and focus on limited types of sequence variations, such as protein point mutations. Thus, extending their extraction scope requires significant manual efforts in examining new instances and developing corresponding rules. As such, new automatic approaches are greatly needed for extracting different kinds of mutations with high accuracy.

RESULTS

Here, we report tmVar, a text-mining approach based on conditional random field (CRF) for extracting a wide range of sequence variants described at protein, DNA and RNA levels according to a standard nomenclature developed by the Human Genome Variation Society. By doing so, we cover several important types of mutations that were not considered in past studies. Using a novel CRF label model and feature set, our method achieves higher performance than a state-of-the-art method on both our corpus (91.4 versus 78.1% in F-measure) and their own gold standard (93.9 versus 89.4% in F-measure). These results suggest that tmVar is a high-performance method for mutation extraction from biomedical literature.

AVAILABILITY

tmVar software and its corpus of 500 manually curated abstracts are available for download at http://www.ncbi.nlm.nih.gov/CBBresearch/Lu/pub/tmVar

Application of text-mining for updating protein post-translational modification annotation in UniProtKB

BACKGROUND

The annotation of protein post-translational modifications (PTMs) is an important task of UniProtKB curators and, with continuing improvements in experimental methodology, an ever greater number of articles are being published on this topic. To help curators cope with this growing body of information we have developed a system which extracts information from the scientific literature for the most frequently annotated PTMs in UniProtKB.

RESULTS

The procedure uses a pattern-matching and rule-based approach to extract sentences with information on the type and site of modification. A ranked list of protein candidates for the modification is also provided. For PTM extraction, precision varies from 57% to 94%, and recall from 75% to 95%, according to the type of modification. The procedure was used to track new publications on PTMs and to recover potential supporting evidence for phosphorylation sites annotated based on the results of large scale proteomics experiments.

CONCLUSIONS

The information retrieval and extraction method we have developed in this study forms the basis of a simple tool for the manual curation of protein post-translational modifications in UniProtKB/Swiss-Prot. Our work demonstrates that even simple text-mining tools can be effectively adapted for database curation tasks, providing that a thorough understanding of the working process and requirements are first obtained. This system can be accessed at http://eagl.unige.ch/PTM/.

Biological network extraction from scientific literature: state of the art and challenges

Networks of molecular interactions explain complex biological processes, and all known information on molecular events is contained in a number of public repositories including the scientific literature. Metabolic and signalling pathways are often viewed separately, even though both types are composed of interactions involving proteins and other chemical entities. It is necessary to be able to combine data from all available resources to judge the functionality, complexity and completeness of any given network overall, but especially the full integration of relevant information from the scientific literature is still an ongoing and complex task. Currently, the text-mining research community is steadily moving towards processing the full body of the scientific literature by making use of rich linguistic features such as full text parsing, to extract biological interactions. The next step will be to combine these with information from scientific databases to support hypothesis generation for the discovery of new knowledge and the extension of biological networks. The generation of comprehensive networks requires technologies such as entity grounding, coordination resolution and co-reference resolution, which are not fully solved and are required to further improve the quality of results. Here, we analyse the state of the art for the extraction of network information from the scientific literature and the evaluation of extraction methods against reference corpora, discuss challenges involved and identify directions for future research.

Gimli: open source and high-performance biomedical name recognition

Background

Automatic recognition of biomedical names is an essential task in biomedical information extraction, presenting several complex and unsolved challenges. In recent years, various solutions have been implemented to tackle this problem. However, limitations regarding system characteristics, customization and usability still hinder their wider application outside text mining research.

Results

We present Gimli, an open-source, state-of-the-art tool for automatic recognition of biomedical names. Gimli includes an extended set of implemented and user-selectable features, such as orthographic, morphological, linguistic-based, conjunctions and dictionary-based. A simple and fast method to combine different trained models is also provided. Gimli achieves an F-measure of 87.17% on GENETAG and 72.23% on JNLPBA corpus, significantly outperforming existing open-source solutions.

Conclusions

Gimli is an off-the-shelf, ready to use tool for named-entity recognition, providing trained and optimized models for recognition of biomedical entities from scientific text. It can be used as a command line tool, offering full functionality, including training of new models and customization of the feature set and model parameters through a configuration file. Advanced users can integrate Gimli in their text mining workflows through the provided library, and extend or adapt its functionalities. Based on the underlying system characteristics and functionality, both for final users and developers, and on the reported performance results, we believe that Gimli is a state-of-the-art solution for biomedical NER, contributing to faster and better research in the field. Gimli is freely available at http://bioinformatics.ua.pt/gimli.

Collaborative biocuration--text-mining development task for document prioritization for curation

The Critical Assessment of Information Extraction systems in Biology (BioCreAtIvE) challenge evaluation is a community-wide effort for evaluating text mining and information extraction systems for the biological domain. The 'BioCreative Workshop 2012' subcommittee identified three areas, or tracks, that comprised independent, but complementary aspects of data curation in which they sought community input: literature triage (Track I); curation workflow (Track II) and text mining/natural language processing (NLP) systems (Track III). Track I participants were invited to develop tools or systems that would effectively triage and prioritize articles for curation and present results in a prototype web interface. Training and test datasets were derived from the Comparative Toxicogenomics Database (CTD; http://ctdbase.org) and consisted of manuscripts from which chemical-gene-disease data were manually curated. A total of seven groups participated in Track I. For the triage component, the effectiveness of participant systems was measured by aggregate gene, disease and chemical 'named-entity recognition' (NER) across articles; the effectiveness of 'information retrieval' (IR) was also measured based on 'mean average precision' (MAP). Top recall scores for gene, disease and chemical NER were 49, 65 and 82%, respectively; the top MAP score was 80%. Each participating group also developed a prototype web interface; these interfaces were evaluated based on functionality and ease-of-use by CTD's biocuration project manager. In this article, we present a detailed description of the challenge and a summary of the results.

SR4GN: a species recognition software tool for gene normalization

As suggested in recent studies, species recognition and disambiguation is one of the most critical and challenging steps in many downstream text-mining applications such as the gene normalization task and protein-protein interaction extraction. We report SR4GN: an open source tool for species recognition and disambiguation in biomedical text. In addition to the species detection function in existing tools, SR4GN is optimized for the Gene Normalization task. As such it is developed to link detected species with corresponding gene mentions in a document. SR4GN achieves 85.42% in accuracy and compares favorably to the other state-of-the-art techniques in benchmark experiments. Finally, SR4GN is implemented as a standalone software tool, thus making it convenient and robust for use in many text-mining applications. SR4GN can be downloaded at: http://www.ncbi.nlm.nih.gov/CBBresearch/Lu/downloads/SR4GN.

Harmonization of gene/protein annotations: towards a gold standard MEDLINE

MOTIVATION

The recognition of named entities (NER) is an elementary task in biomedical text mining. A number of NER solutions have been proposed in recent years, taking advantage of available annotated corpora, terminological resources and machine-learning techniques. Currently, the best performing solutions combine the outputs from selected annotation solutions measured against a single corpus. However, little effort has been spent on a systematic analysis of methods harmonizing the annotation results and measuring against a combination of Gold Standard Corpora (GSCs).

RESULTS

We present Totum, a machine learning solution that harmonizes gene/protein annotations provided by heterogeneous NER solutions. It has been optimized and measured against a combination of manually curated GSCs. The performed experiments show that our approach improves the F-measure of state-of-the-art solutions by up to 10% (achieving ≈70%) in exact alignment and 22% (achieving ≈82%) in nested alignment. We demonstrate that our solution delivers reliable annotation results across the GSCs and it is an important contribution towards a homogeneous annotation of MEDLINE abstracts.

AVAILABILITY AND IMPLEMENTATION

Totum is implemented in Java and its resources are available at http://bioinformatics.ua.pt/totum

Boosting performance of gene mention tagging system by hybrid methods

NER (Named Entity Recognition) in biomedical literature is presently one of the internationally concerned NLP (Natural Language Processing) research questions. In order to get higher performance, a hybrid experimental framework is presented for the gene mention tagging task. Six classifiers are firstly constructed by four toolkits (CRF++, YamCha, Maximum Entropy (ME) and MALLET) with different training methods and features sets, and then combined with three different hybrid methods respectively: simple set operation method, voting method and two layer stacking method. Experiments carried out on the corpus of BioCreative II GM task show that the three hybrid methods get the F-measure of 87.40%, 87.31% and 87.70% separately without any post-processing, which are all higher than those of any single ones. Our best hybrid method (two layer stacking method) achieves an F-measure of 88.42% after post-processing, which outperforms most of the state-of-the-art systems. We also discuss the influence on the performance of the ensemble system by the number, performance and divergence of single classifiers in each hybrid method, and give the corresponding analysis why our hybrid models can improve the performance.

Soft tagging of overlapping high confidence gene mention variants for cross-species full-text gene normalization

Background

Previously, gene normalization (GN) systems are mostly focused on disambiguation using contextual information. An effective gene mention tagger is deemed unnecessary because the subsequent steps will filter out false positives and high recall is sufficient. However, unlike similar tasks in the past BioCreative challenges, the BioCreative III GN task is particularly challenging because it is not species-specific. Required to process full-length articles, an ineffective gene mention tagger may produce a huge number of ambiguous false positives that overwhelm subsequent filtering steps while still missing many true positives.

Results

We present our GN system participated in the BioCreative III GN task. Our system applies a typical 2-stage approach to GN but features a soft tagging gene mention tagger that generates a set of overlapping gene mention variants with a nearly perfect recall. The overlapping gene mention variants increase the chance of precise match in the dictionary and alleviate the need of disambiguation. Our GN system achieved a precision of 0.9 (F-score 0.63) on the BioCreative III GN test corpus with the silver annotation of 507 articles. Its TAP-k scores are competitive to the best results among all participants.

Conclusions

We show that despite the lack of clever disambiguation in our gene normalization system, effective soft tagging of gene mention variants can indeed contribute to performance in cross-species and full-text gene normalization.

Cross-species gene normalization by species inference

Background

To access and utilize the rich information contained in the biomedical literature, the ability to recognize and normalize gene mentions referenced in the literature is crucial. In this paper, we focus on improvements to the accuracy of gene normalization in cases where species information is not provided. Gene names are often ambiguous, in that they can refer to the genes of many species. Therefore, gene normalization is a difficult challenge.

Methods

We define “gene normalization” as a series of tasks involving several issues, including gene name recognition, species assignation and species-specific gene normalization. We propose an integrated method, GenNorm, consisting of three modules to handle the issues of this task. Every issue can affect overall performance, though the most important is species assignation. Clearly, correct identification of the species can decrease the ambiguity of orthologous genes.

Results

In experiments, the proposed model attained the top-1 threshold average precision (TAP-k) scores of 0.3297 (k=5), 0.3538 (k=10), and 0.3535 (k=20) when tested against 50 articles that had been selected for their difficulty and the most divergent results from pooled team submissions. In the silver-standard-507 evaluation, our TAP-k scores are 0.4591 for k=5, 10, and 20 and were ranked 2^nd, 2^nd, and 3^rd respectively.

Availability

A web service and input, output formats of GenNorm are available at http://ikmbio.csie.ncku.edu.tw/GN/.

The gene normalization task in BioCreative III

BACKGROUND

We report the Gene Normalization (GN) challenge in BioCreative III where participating teams were asked to return a ranked list of identifiers of the genes detected in full-text articles. For training, 32 fully and 500 partially annotated articles were prepared. A total of 507 articles were selected as the test set. Due to the high annotation cost, it was not feasible to obtain gold-standard human annotations for all test articles. Instead, we developed an Expectation Maximization (EM) algorithm approach for choosing a small number of test articles for manual annotation that were most capable of differentiating team performance. Moreover, the same algorithm was subsequently used for inferring ground truth based solely on team submissions. We report team performance on both gold standard and inferred ground truth using a newly proposed metric called Threshold Average Precision (TAP-k).

RESULTS

We received a total of 37 runs from 14 different teams for the task. When evaluated using the gold-standard annotations of the 50 articles, the highest TAP-k scores were 0.3297 (k=5), 0.3538 (k=10), and 0.3535 (k=20), respectively. Higher TAP-k scores of 0.4916 (k=5, 10, 20) were observed when evaluated using the inferred ground truth over the full test set. When combining team results using machine learning, the best composite system achieved TAP-k scores of 0.3707 (k=5), 0.4311 (k=10), and 0.4477 (k=20) on the gold standard, representing improvements of 12.4%, 21.8%, and 26.6% over the best team results, respectively.

CONCLUSIONS

By using full text and being species non-specific, the GN task in BioCreative III has moved closer to a real literature curation task than similar tasks in the past and presents additional challenges for the text mining community, as revealed in the overall team results. By evaluating teams using the gold standard, we show that the EM algorithm allows team submissions to be differentiated while keeping the manual annotation effort feasible. Using the inferred ground truth we show measures of comparative performance between teams. Finally, by comparing team rankings on gold standard vs. inferred ground truth, we further demonstrate that the inferred ground truth is as effective as the gold standard for detecting good team performance.

Recent progress in automatically extracting information from the pharmacogenomic literature

The biomedical literature holds our understanding of pharmacogenomics, but it is dispersed across many journals. In order to integrate our knowledge, connect important facts across publications and generate new hypotheses we must organize and encode the contents of the literature. By creating databases of structured pharmocogenomic knowledge, we can make the value of the literature much greater than the sum of the individual reports. We can, for example, generate candidate gene lists or interpret surprising hits in genome-wide association studies. Text mining automatically adds structure to the unstructured knowledge embedded in millions of publications, and recent years have seen a surge in work on biomedical text mining, some specific to pharmacogenomics literature. These methods enable extraction of specific types of information and can also provide answers to general, systemic queries. In this article, we describe the main tasks of text mining in the context of pharmacogenomics, summarize recent applications and anticipate the next phase of text mining applications.

Using machine learning for concept extraction on clinical documents from multiple data sources

OBJECTIVE

Concept extraction is a process to identify phrases referring to concepts of interests in unstructured text. It is a critical component in automated text processing. We investigate the performance of machine learning taggers for clinical concept extraction, particularly the portability of taggers across documents from multiple data sources.

METHODS

We used BioTagger-GM to train machine learning taggers, which we originally developed for the detection of gene/protein names in the biology domain. Trained taggers were evaluated using the annotated clinical documents made available in the 2010 i2b2/VA Challenge workshop, consisting of documents from four data sources.

RESULTS

As expected, performance of a tagger trained on one data source degraded when evaluated on another source, but the degradation of the performance varied depending on data sources. A tagger trained on multiple data sources was robust, and it achieved an F score as high as 0.890 on one data source. The results also suggest that performance of machine learning taggers is likely to improve if more annotated documents are available for training.

CONCLUSION

Our study shows how the performance of machine learning taggers is degraded when they are ported across clinical documents from different sources. The portability of taggers can be enhanced by training on datasets from multiple sources. The study also shows that BioTagger-GM can be easily extended to detect clinical concept mentions with good performance.

A framework for semisupervised feature generation and its applications in biomedical literature mining

Feature representation is essential to machine learning and text mining. In this paper, we present a feature coupling generalization (FCG) framework for generating new features from unlabeled data. It selects two special types of features, i.e., example-distinguishing features (EDFs) and class-distinguishing features (CDFs) from original feature set, and then generalizes EDFs into higher-level features based on their coupling degrees with CDFs in unlabeled data. The advantage is: EDFs with extreme sparsity in labeled data can be enriched by their co-occurrences with CDFs in unlabeled data so that the performance of these low-frequency features can be greatly boosted and new information from unlabeled can be incorporated. We apply this approach to three tasks in biomedical literature mining: gene named entity recognition (NER), protein-protein interaction extraction (PPIE), and text classification (TC) for gene ontology (GO) annotation. New features are generated from over 20 GB unlabeled PubMed abstracts. The experimental results on BioCreative 2, AIMED corpus, and TREC 2005 Genomics Track show that 1) FCG can utilize well the sparse features ignored by supervised learning. 2) It improves the performance of supervised baselines by 7.8 percent, 5.0 percent, and 5.8 percent, respectively, in the tree tasks. 3) Our methods achieve 89.1, 64.5 F-score, and 60.1 normalized utility on the three benchmark data sets.

eFIP: a tool for mining functional impact of phosphorylation from literature

Technologies and experimental strategies have improved dramatically in the field of genomics and proteomics facilitating analysis of cellular and biochemical processes, as well as of proteins networks. Based on numerous such analyses, there has been a significant increase of publications in life sciences and biomedicine. In this respect, knowledge bases are struggling to cope with the literature volume and they may not be able to capture in detail certain aspects of proteins and genes. One important aspect of proteins is their phosphorylated states and their implication in protein function and protein interacting networks. For this reason, we developed eFIP, a web-based tool, which aids scientists to find quickly abstracts mentioning phosphorylation of a given protein (including site and kinase), coupled with mentions of interactions and functional aspects of the protein. eFIP combines information provided by applications such as eGRAB, RLIMS-P, eGIFT and AIIAGMT, to rank abstracts mentioning phosphorylation, and to display the results in a highlighted and tabular format for a quick inspection. In this chapter, we present a case study of results returned by eFIP for the protein BAD, which is a key regulator of apoptosis that is posttranslationally modified by phosphorylation.

From NPC therapeutic target identification to potential treatment strategy

Nasopharyngeal carcinoma (NPC) is relatively rare in Western countries but is a common cancer in southern Asia. Many differentially expressed genes have been linked to NPC; however, how to prioritize therapeutic targets and potential drugs from unsorted gene lists remains largely unknown. We first collected 558 upregulated and 993 downregulated NPC genes from published microarray data and the primary literatures. We then postulated that conversion of gene signatures into the protein-protein interaction network and analyzing the network topologically could provide insight into key regulators involved in tumorigenesis of NPC. Of particular interest was the presence of cliques, called fully connected subgraphs, in the inferred NPC networks. These clique-based hubs, connecting with more than three queries and ranked higher than other nodes in the NPC protein-protein interaction network, were further narrowed down by pathway analysis to retrieve 24 upregulated and 6 downregulated bottleneck genes for predicting NPC carcinogenesis. Moreover, additional oncogenes, tumor suppressor genes, genes involved in protein complexes, and genes obtained after functional profiling were merged with the bottleneck genes to form the final gene signature of 38 upregulated and 10 downregulated genes. We used the initial and final NPC gene signatures to query the Connectivity Map, respectively, and found that target reduction through our pipeline could efficiently uncover potential drugs with cytotoxicity to NPC cancer cells. An integrative Web site (http://140.109.23.188:8080/NPC) was established to facilitate future NPC research. This in silico approach, from target prioritization to potential drugs identification, might be an effective method for various cancer researches.

BIOADI: a machine learning approach to identifying abbreviations and definitions in biological literature

BACKGROUND

To automatically process large quantities of biological literature for knowledge discovery and information curation, text mining tools are becoming essential. Abbreviation recognition is related to NER and can be considered as a pair recognition task of a terminology and its corresponding abbreviation from free text. The successful identification of abbreviation and its corresponding definition is not only a prerequisite to index terms of text databases to produce articles of related interests, but also a building block to improve existing gene mention tagging and gene normalization tools.

RESULTS

Our approach to abbreviation recognition (AR) is based on machine-learning, which exploits a novel set of rich features to learn rules from training data. Tested on the AB3P corpus, our system demonstrated a F-score of 89.90% with 95.86% precision at 84.64% recall, higher than the result achieved by the existing best AR performance system. We also annotated a new corpus of 1200 PubMed abstracts which was derived from BioCreative II gene normalization corpus. On our annotated corpus, our system achieved a F-score of 86.20% with 93.52% precision at 79.95% recall, which also outperforms all tested systems.

CONCLUSION

By applying our system to extract all short form-long form pairs from all available PubMed abstracts, we have constructed BIOADI. Mining BIOADI reveals many interesting trends of bio-medical research. Besides, we also provide an off-line AR software in the download section on http://bioagent.iis.sinica.edu.tw/BIOADI/.