RESUMEN
Iodine-129, which is a promising tracer for dating old groundwater, has been used as a tracer for deep upwelling groundwater. The nuclide is expected to be one of the key factors for site selection for high-level radioactive waste disposal, which is a global societal issue. The pre-anthropogenic 129I/127I ratio for marine iodine is (1.50 ± 0.15) × 10-12, which could be considered the initial value for 129I dating. This study identifies the challenges in groundwater age dating using 129I/127I. We measured the ratios of 129I/127I and 81Kr/Kr and concentration of 4He in groundwater from boreholes on the northern coast of Japan. The 129I dating results were not coincident with the other groundwater dating results. The iodine in the groundwater was inferred to be released in situ from marine organisms in sediments of various ages. We estimated that the primordial iodine ratio originating from seawater was ~ 1 × 10-13 (8 × 10-14 ~ 2 × 10-13). The groundwater age deduced from the 129I/127I ratio using this value agrees with other groundwater dating results.
RESUMEN
We investigated the tritium concentration in commercial modern D2O reagents frequently used in nuclear magnetic resonance analysis for analytical chemistry and in environmental tracer testing. The concentration of tritium in 11 D2O and 1 H218O reagents ranged from 61 Bq/L (5 × 102 TU) to 2.5 × 103 Bq/L (2 × 104 TU) in order of magnitude. The tritium concentration in the D2O reagents have increased with the increasing purity of D2O. The tritium concentration in all reagents was an order of magnitude greater than that in the surface waters at the Fukushima off-site of the Fukushima Daiichi Nuclear Power Plant after the accident in 2011 and in precipitation during the nuclear test era. However, the concentration of the tritium was lower than the regulatory limit for the concentration of tritium in drinking water accepted by the World Health Organization guidelines. The internal exposure effects from drinking the tritium water, which is contaminated by the tritium condensed in the reagent production processes, were negligible, even if the reagent was used in the environmental tracer test.
RESUMEN
The opportunity to measure the concentrations of 3H and 36Cl released by the Fukushima nuclear accident in 2011 directly in rain was lost in the early stage of the accident. We have, however, been able to reconstruct the deposition record of atmospheric 3H and 36Cl following the accident using a bore hole that was drilled in 2014 at Koriyama at a distance of 60 km from the accident. The contributions of 3H and 36Cl from the accident are 1.4 × 1013 and 2.0 × 1012 atoms m-2 respectively at this site. Very high concentrations of both 3H (46 Bq L-1) and 36Cl (3.36 × 1011 atoms L-1) were found in the unsaturated soil at depths between 300 and 350 cm. From these, conservative estimates for the 3H and 36Cl concentrations in the precipitation in the ~ 6 weeks following the accident were 607 Bq L-1 and 4.74 × 1010 atoms L-1, respectively. A second hole drilled in 2016 showed that 3H concentrations in the unsaturated soil and shallow groundwater had returned to close to natural levels, although the 36Cl concentrations were still significantly elevated above natural levels.
RESUMEN
MOTIVATION: The recognition of mentions of species names in text is a critically important task for biomedical text mining. While deep learning-based methods have made great advances in many named entity recognition tasks, results for species name recognition remain poor. We hypothesize that this is primarily due to the lack of appropriate corpora. RESULTS: We introduce the S1000 corpus, a comprehensive manual re-annotation and extension of the S800 corpus. We demonstrate that S1000 makes highly accurate recognition of species names possible (F-score =93.1%), both for deep learning and dictionary-based methods. AVAILABILITY AND IMPLEMENTATION: All resources introduced in this study are available under open licenses from https://jensenlab.org/resources/s1000/. The webpage contains links to a Zenodo project and three GitHub repositories associated with the study.
Asunto(s)
Minería de Datos , Minería de Datos/métodosRESUMEN
Chaperone-mediated autophagy (CMA) is a pathway in the autophagy-lysosome protein degradation system. CMA impairment has been implicated to play a role in spinocerebellar ataxia (SCA) pathogenesis. D-cysteine is metabolized by D-amino acid oxidase (DAO), leading to hydrogen sulfide generation in the cerebellum. Although D-cysteine alleviates the disease phenotypes in SCA-model mice, it remains unknown how hydrogen sulfide derived from D-cysteine exerts this effect. In the present study, we investigated the effects of D-cysteine and hydrogen sulfide on CMA activity using a CMA activity marker that we have established. D-cysteine activated CMA in Purkinje cells (PCs) of primary cerebellar cultures where DAO was expressed, while it failed to activate CMA in DAO-deficient AD293 cells. In contrast, Na2S, a hydrogen sulfide donor, activated CMA in both PCs and AD293 cells. Nuclear factor erythroid 2-related factor 2 (Nrf2) is known to be activated by hydrogen sulfide and regulate CMA activity. An Nrf2 inhibitor, ML385, prevented CMA activation triggered by D-cysteine and Na2S. Additionally, long-term treatment with D-cysteine increased the amounts of Nrf2 and LAMP2A, a CMA-related protein, in the mouse cerebellum. These findings suggest that hydrogen sulfide derived from D-cysteine enhances CMA activity via Nrf2 activation.
Asunto(s)
Autofagia Mediada por Chaperones , Sulfuro de Hidrógeno , Animales , Ratones , Cisteína/metabolismo , Cisteína/farmacología , Sulfuro de Hidrógeno/metabolismo , Sulfuro de Hidrógeno/farmacología , Factor 2 Relacionado con NF-E2/metabolismo , Células de PurkinjeRESUMEN
Spinocerebellar ataxia (SCA) is a group of autosomal-dominantly inherited ataxia and is classified into SCA1-48 by the difference of causal genes. Several SCA-causing proteins commonly impair dendritic development in primary cultured Purkinje cells (PCs). We assume that primary cultured PCs expressing SCA-causing proteins are available as in vitro SCA models and that chemicals that improve the impaired dendritic development would be effective for various SCAs. We have recently revealed that D-cysteine enhances the dendritic growth of primary cultured PCs via hydrogen sulfide production. In the present study, we first investigated whether D-cysteine is effective for in vitro SCA models. We expressed SCA1-, SCA3-, and SCA21-causing mutant proteins to primary cultured PCs using adeno-associated viral serotype 9 (AAV9) vectors. D-Cysteine (0.2 mM) significantly ameliorated the impaired dendritic development commonly observed in primary cultured PCs expressing these three SCA-causing proteins. Next, we investigated the therapeutic effect of long-term treatment with D-cysteine on an in vivo SCA model. SCA1 model mice were established by the cerebellar injection of AAV9 vectors, which express SCA1-causing mutant ataxin-1, to ICR mice. Long-term treatment with D-cysteine (100 mg/kg/day) significantly inhibited the progression of motor dysfunction in SCA1 model mice. Immunostaining experiments revealed that D-cysteine prevented the reduction of mGluR1 and glial activation at the early stage after the onset of motor dysfunction in SCA1 model mice. These findings strongly suggest that D-cysteine has therapeutic potential against in vitro and in vivo SCA models and may be a novel therapeutic agent for various SCAs.
Asunto(s)
Ataxina-1/biosíntesis , Cerebelo/efectos de los fármacos , Cerebelo/metabolismo , Cisteína/uso terapéutico , Ataxias Espinocerebelosas/tratamiento farmacológico , Ataxias Espinocerebelosas/metabolismo , Animales , Ataxina-1/genética , Células Cultivadas , Cisteína/farmacología , Femenino , Masculino , Ratones , Ratones Endogámicos ICR , Ratones Transgénicos , Embarazo , Ratas , Ratas Wistar , Ataxias Espinocerebelosas/genéticaRESUMEN
Although 10 years have passed since Japan's Fukushima nuclear accident, the future radiation risk from 137Cs contamination of wood via root uptake is a serious concern. We estimated the depth at which the roots of evergreen coniferous sugi (Cryptomeria japonica) and broadleaf deciduous konara (Quercus serrata) trees actively take up soil water by using positive δD values from the artificial D2O tracer and seasonal changes in the δ18O values of soil water as a natural environmental tracer. We compared the tracer concentration changes in xylem sap with those in the soil water and ascertained that both tree species primarily took up water from a depth of 20 cm, though with mixing of water from other depths. Using sap hydrodynamics in tree stems, we found that water circulation was significantly slower in heartwood than in sapwood. Heartwood water was not supplied by direct root uptake of soil water. The measured diffusion coefficients for D2O, K+, Cs+, and I- in xylem stems were greater in sapwood than in heartwood, and their magnitude was inversely correlated with their molecular weights. The distribution of D2O and 137Cs concentrations along the radial stem could be explained by simulations using the simple advective diffusion model.
RESUMEN
BACKGROUND AND PURPOSE: Niemann-Pick disease type C (NPC) is a lysosomal storage disorder with disrupted intracellular cholesterol trafficking. A cyclic heptasaccharide, 2-hydroxypropyl-ß-cyclodextrin (HP-ß-CD), is a cholesterol solubilizer that is being developed to treat NPC, but its ototoxicity and pulmonary toxicity remain important issues. We have characterized 2-hydroxypropyl-γ-cyclodextrin (HP-γ-CD), a cyclic octasaccharide with a larger cavity than HP-ß-CD, as a candidate drug to treat NPC. However, the molecular target of HP-γ-CD with respect to NPC and its potential for clinical application are still unclear. EXPERIMENTAL APPROACH: We investigated the mode of interaction between HP-γ-CD and cholesterol by phase-solubility analysis, proton NMR spectroscopy and molecular dynamics simulations. We then evaluated the therapeutic effects of HP-γ-CD compared with HP-ß-CD using cellular and murine NPC models. Mouse auditory and pulmonary function tests were also conducted. KEY RESULTS: HP-γ-CD solely formed a 1:1 inclusion complex with cholesterol with an affinity similar to that of HP-ß-CD. In vitro, HP-γ-CD and HP-ß-CD amelioration of NPC-related manifestations was almost equivalent at lower concentrations. However, at higher concentrations, the cholesterol inclusion mode of HP-ß-CD shifted to the highly soluble 2:1 complex whereas that of HP-γ-CD maintained solely the 1:1 complex. The constant lower cholesterol solubilizing ability of HP-γ-CD conferred it with significantly reduced toxicity compared with HP-ß-CD, but equal efficacy in treating a mouse model of NPC. CONCLUSIONS AND IMPLICATIONS: HP-γ-CD can serve as a fine-tuned cholesterol solubilizer for the treatment of NPC with a wider safety margin than HP-ß-CD in terms of ototoxicity and pulmonary toxicity.
Asunto(s)
Ciclodextrinas , Enfermedad de Niemann-Pick Tipo C , 2-Hidroxipropil-beta-Ciclodextrina , Animales , Colesterol , Modelos Animales de Enfermedad , Ratones , Enfermedad de Niemann-Pick Tipo C/tratamiento farmacológicoRESUMEN
AIMS: Chaperone-mediated autophagy (CMA) is a pathway involved in the autophagy lysosome protein degradation system. CMA has attracted attention as a contributing factor to neurodegenerative diseases since it participates in the degradation of disease-causing proteins. We previously showed that CMA is generally impaired in cells expressing the proteins causing spinocerebellar ataxias (SCAs). Therefore, we investigated the effect of CMA impairment on motor function and the neural survival of cerebellar neurons using the micro RNA (miRNA)-mediated knockdown of lysosome-associated protein 2A (LAMP2A), a CMA-related protein. METHODS: We injected adeno-associated virus serotype 9 vectors, which express green fluorescent protein (GFP) and miRNA (negative control miRNA or LAMP2A miRNA) under neuron-specific synapsin I promoter, into cerebellar parenchyma of 4-week-old ICR mice. Motor function of mice was evaluated by beam walking and footprint tests. Immunofluorescence experiments of cerebellar slices were conducted to evaluate histological changes in cerebella. RESULTS: GFP and miRNA were expressed in interneurons (satellite cells and basket cells) in molecular layers and granule cells in the cerebellar cortices, but not in cerebellar Purkinje cells. LAMP2A knockdown in cerebellar neurons triggered progressive motor impairment, prominent loss of cerebellar Purkinje cells, interneurons, granule cells at the late stage, and astrogliosis and microgliosis from the early stage. CONCLUSIONS: CMA impairment in cerebellar interneurons and granule cells triggers the progressive ataxic phenotype, gliosis and the subsequent degeneration of cerebellar neurons, including Purkinje cells. Our present findings strongly suggest that CMA impairment is related to the pathogenesis of various SCAs.
Asunto(s)
Ataxia Cerebelosa/patología , Cerebelo/patología , Autofagia Mediada por Chaperones/fisiología , Proteína 2 de la Membrana Asociada a los Lisosomas/metabolismo , Degeneración Nerviosa/patología , Neuronas/patología , Animales , Ataxia Cerebelosa/metabolismo , Cerebelo/metabolismo , Ratones Endogámicos ICR , Degeneración Nerviosa/metabolismo , Neuronas/metabolismo , FenotipoRESUMEN
Sirtuin interacts with many regulatory proteins involved in energy homeostasis, DNA repair, cell survival, and lifespan extension. We investigated the functional roles of Sir2D during early Dictyostelium development upon starvation. We found that ectopic expression of Sir2D accelerated development among three Sirtuins containing highly homologous catalytic domain sequences to mouse Sirt1. Sir2D expression upregulated adenylate cyclase A (aca) mRNA expression 2, 4 and 6 h after starvation. We have previously reported that nicotinamide, a Sirt1 inhibitor, treatment delayed the development and decreased the expression of aca at 4 h after starvation. Sir2D expressing cells showed resistance against the nicotinamide effect. RNAi-mediated Sir2D knockdown cells were generated, and their development was also delayed. Aca expression was decreased 4 h after starvation. Sir2D expression restored the developmental impairment of Sir2D knockdown cells. The induction of aca upon starvation starts with transcriptional activation of MybB. The ectopic expression of MybB accelerated the development and increased the expression of aca 2 and 4 h after starvation but did not restore the phenotype of Sir2D knockdown cells. Sir2D expression had no effects on MybB-null mutant cells during early development. Thus, MybB is necessary for the upregulation of aca by Sir2D, and Sir2D is necessary for the full induction of aca after 4 h by MybB. MybB was coimmunoprecipitated with Sir2D, suggesting an interaction between MybB and Sir2D. These results suggest that Sir2D regulates aca expression through interaction with the MybB transcription factor early in Dictyostelium development upon starvation.
RESUMEN
Spinocerebellar ataxia type 21 (SCA21) is caused by missense or nonsense mutations of the transmembrane protein 240 (TMEM240). Molecular mechanisms of SCA21 pathogenesis remain unknown because the functions of TMEM240 have not been elucidated. We aimed to reveal the molecular pathogenesis of SCA21 using cell and mouse models that overexpressed the wild-type and SCA21 mutant TMEM240. In HeLa cells, overexpressed TMEM240 localized around large cytoplasmic vesicles. The SCA21 mutation did not affect this localization. Because these vesicles contained endosomal markers, we evaluated the effect of TMEM240 fused with a FLAG tag (TMEM-FL) on endocytosis and autophagic protein degradation. Wild-type TMEM-FL significantly impaired clathrin-mediated endocytosis, whereas the SCA21 mutants did not. The SCA21 mutant TMEM-FL significantly impaired autophagic lysosomal protein degradation, in contrast to wild-type. Next, we investigated how TMEM240 affects the neural morphology of primary cultured cerebellar Purkinje cells (PCs). The SCA21 mutant TMEM-FL significantly prevented the dendritic development of PCs, in contrast to the wild-type. Finally, we assessed mice that expressed wild-type or SCA21 mutant TMEM-FL in cerebellar neurons using adeno-associated viral vectors. Mice expressing the SCA21 mutant TMEM-FL showed impaired motor coordination. Although the SCA21 mutant TMEM-FL did not trigger neurodegeneration, activation of microglia and astrocytes was induced before motor miscoordination. In addition, immunoblot experiments revealed that autophagic lysosomal protein degradation, especially chaperone-mediated autophagy, was also impaired in the cerebella that expressed the SCA21 mutant TMEM-FL. These dysregulated functions in vitro, and induction of early gliosis and lysosomal impairment in vivo by the SCA21 mutant TMEM240 may contribute to the pathogenesis of SCA21.
Asunto(s)
Lisosomas/metabolismo , Proteínas de la Membrana/biosíntesis , Mutación/fisiología , Neuroglía/metabolismo , Degeneraciones Espinocerebelosas/metabolismo , Animales , Femenino , Células HeLa , Humanos , Lisosomas/genética , Lisosomas/patología , Masculino , Proteínas de la Membrana/genética , Ratones , Ratones Endogámicos C57BL , Ratones Transgénicos , Neuroglía/patología , Embarazo , Ratas , Ratas Wistar , Degeneraciones Espinocerebelosas/genéticaRESUMEN
Magnaporthe oryzae chrysovirus 1 strain A (MoCV1-A) is the causal agent of growth repression and attenuated virulence (hypovirulence) of the rice blast fungus, Magnaporthe oryzae. We previously revealed that heterologous expression of the MoCV1-A ORF4 protein resulted in cytological damage to the yeasts Saccharomyces cerevisiae and Cryptococcus neoformans. Since the ORF4 protein is one of the components of viral particles, we evaluated the inhibitory effects of the purified virus particle against the conidial germination of M. oryzae, and confirmed its suppressive effects. Recombinant MoCV1-A ORF4 protein produced in Pichia pastoris was also effective for suppression of conidial germination of M. oryzae. MoCV1-A ORF4 protein sequence showed significant similarity to 6 related mycoviral proteins; Botrysphaeria dothidea chrysovirus 1, two Fusarium graminearum viruses, Fusarium oxysporum f. sp. dianthi mycovirus 1, Penicillium janczewski chrysovirus and Agaricus bisporus virus 1 in the Chrysoviridae family. Multiple alignments of the ORF4-related protein sequences showed that their central regions (210-591 aa in MoCV1-A ORF4) are relatively conserved. Indeed, yeast transformants expressing the conserved central region of MoCV1-A ORF4 protein (325-575 aa) showed similar impaired growth phenotypes as those observed in yeasts expressing the full-length MoCV1-A ORF4 protein. These data suggest that the mycovirus itself and its encoded viral protein can be useful as anti-fungal proteins to control rice blast disease caused by M. oryzae and other pathogenic fungi.
Asunto(s)
Virus Fúngicos/fisiología , Germinación , Oryza/crecimiento & desarrollo , Oryza/virología , Virus ARN/fisiología , Proteínas Virales/genética , Proteínas Virales/metabolismo , Secuencia de Aminoácidos , Productos Biológicos , Expresión Génica , Germinación/efectos de los fármacos , Dominios Proteicos , Proteínas Recombinantes/química , Proteínas Recombinantes/genética , Proteínas Recombinantes/metabolismo , Proteínas Recombinantes/farmacología , Saccharomyces cerevisiae/genética , Saccharomyces cerevisiae/metabolismo , Proteínas Virales/química , Proteínas Virales/farmacologíaRESUMEN
MOTIVATION: The recognition and normalization of cell line names in text is an important task in biomedical text mining research, facilitating for instance the identification of synthetically lethal genes from the literature. While several tools have previously been developed to address cell line recognition, it is unclear whether available systems can perform sufficiently well in realistic and broad-coverage applications such as extracting synthetically lethal genes from the cancer literature. In this study, we revisit the cell line name recognition task, evaluating both available systems and newly introduced methods on various resources to obtain a reliable tagger not tied to any specific subdomain. In support of this task, we introduce two text collections manually annotated for cell line names: the broad-coverage corpus Gellus and CLL, a focused target domain corpus. RESULTS: We find that the best performance is achieved using NERsuite, a machine learning system based on Conditional Random Fields, trained on the Gellus corpus and supported with a dictionary of cell line names. The system achieves an F-score of 88.46% on the test set of Gellus and 85.98% on the independently annotated CLL corpus. It was further applied at large scale to 24 302 102 unannotated articles, resulting in the identification of 5 181 342 cell line mentions, normalized to 11 755 unique cell line database identifiers. AVAILABILITY AND IMPLEMENTATION: The manually annotated datasets, the cell line dictionary, derived corpora, NERsuite models and the results of the large-scale run on unannotated texts are available under open licenses at http://turkunlp.github.io/Cell-line-recognition/. CONTACT: sukaew@utu.fi.
Asunto(s)
Minería de Datos/métodos , Bases de Datos Factuales , Genes Letales , Neoplasias/patología , Terminología como Asunto , Inteligencia Artificial , Línea Celular , Biología Computacional/métodos , Humanos , Almacenamiento y Recuperación de la Información , Aprendizaje Automático , Neoplasias/genética , Publicaciones , SemánticaRESUMEN
BACKGROUND: Since their introduction in 2009, the BioNLP Shared Task events have been instrumental in advancing the development of methods and resources for the automatic extraction of information from the biomedical literature. In this paper, we present the Cancer Genetics (CG) and Pathway Curation (PC) tasks, two event extraction tasks introduced in the BioNLP Shared Task 2013. The CG task focuses on cancer, emphasizing the extraction of physiological and pathological processes at various levels of biological organization, and the PC task targets reactions relevant to the development of biomolecular pathway models, defining its extraction targets on the basis of established pathway representations and ontologies. RESULTS: Six groups participated in the CG task and two groups in the PC task, together applying a wide range of extraction approaches including both established state-of-the-art systems and newly introduced extraction methods. The best-performing systems achieved F-scores of 55% on the CG task and 53% on the PC task, demonstrating a level of performance comparable to the best results achieved in similar previously proposed tasks. CONCLUSIONS: The results indicate that existing event extraction technology can generalize to meet the novel challenges represented by the CG and PC task settings, suggesting that extraction methods are capable of supporting the construction of knowledge bases on the molecular mechanisms of cancer and the curation of biomolecular pathway models. The CG and PC tasks continue as open challenges for all interested parties, with data, tools and resources available from the shared task homepage.
Asunto(s)
Redes Reguladoras de Genes , Genes , Almacenamiento y Recuperación de la Información , Bases del Conocimiento , Modelos Teóricos , Neoplasias/genética , Neoplasias/patología , Humanos , Procesamiento de Lenguaje NaturalRESUMEN
Large areas of forests were radioactively contaminated by the Fukushima nuclear accident of 2011, and forest decontamination is now an important problem in Japan. However, whether trees absorb radioactive fallout from soil via the roots or directly from the atmosphere through the bark and leaves is unclear. We measured the uptake of radiocesium by trees in forests heavily contaminated by the Fukushima nuclear accident. The radiocesium concentrations in sapwood of two tree species, the deciduous broadleaved konara (Quercus serrata) and the evergreen coniferous sugi (Cryptomeria japonica), were higher than that in heartwood. The concentration profiles showed anomalous directionality in konara and non-directionality in sugi, indicating that most radiocesium in the tree rings was directly absorbed from the atmosphere via bark and leaves rather than via roots. Numerical modelling shows that the maximum (137)Cs concentration in the xylem of konara will be achieved 28 years after the accident. Conversely, the values for sugi will monotonously decrease because of the small transfer factor in this species. Overall, xylem (137)Cs concentrations will not be affected by root uptake if active root systems occur 10â cm below the soil.
Asunto(s)
Cryptomeria/metabolismo , Accidente Nuclear de Fukushima , Corteza de la Planta/metabolismo , Hojas de la Planta/metabolismo , Quercus/metabolismo , Ceniza Radiactiva , Transporte Biológico , Radioisótopos de Cesio/metabolismo , Cryptomeria/efectos de la radiación , Bosques , Japón , Hojas de la Planta/efectos de la radiación , Raíces de Plantas/metabolismo , Raíces de Plantas/efectos de la radiación , Quercus/efectos de la radiación , Monitoreo de Radiación , Suelo/química , Contaminantes Radiactivos del Suelo/metabolismo , Xilema/metabolismo , Xilema/efectos de la radiaciónRESUMEN
We measured (134)Cs and (137)Cs in the surface soil of the Kanto loam in the eastern Tokyo metropolitan area and the Nishiyama loam in Nagasaki, Japan. The observed (137)Cs deposition in the Kanto loam from the Fukushima nuclear power plant (NPP) accident ranged from 4.0 to 77 kBq m(-2), which corresponds to 0.3-5 times of that in the Nishiyama loam. The (137)Cs retardation factor in the Kanto loam obtained seven months after the Fukusima NPP accident and in the Nishiyama loam after 36 and 38 years from the detonation of the Pu atomic bomb (A-bomb) ranged from 180 to 260 and 2000 to 10,000, respectively. This difference in the retardation factors is attributed to an aging effect that corresponds to seven months and 36 to 38 years after the deposition of (137)Cs occurred on the soil minerals.
Asunto(s)
Radioisótopos de Cesio/análisis , Explosiones , Accidente Nuclear de Fukushima , Armas Nucleares , Suelo/química , Adsorción , Radioisótopos de Cesio/química , Japón , Propiedades de Superficie , Factores de TiempoRESUMEN
MOTIVATION: To create, verify and maintain pathway models, curators must discover and assess knowledge distributed over the vast body of biological literature. Methods supporting these tasks must understand both the pathway model representations and the natural language in the literature. These methods should identify and order documents by relevance to any given pathway reaction. No existing system has addressed all aspects of this challenge. METHOD: We present novel methods for associating pathway model reactions with relevant publications. Our approach extracts the reactions directly from the models and then turns them into queries for three text mining-based MEDLINE literature search systems. These queries are executed, and the resulting documents are combined and ranked according to their relevance to the reactions of interest. We manually annotate document-reaction pairs with the relevance of the document to the reaction and use this annotation to study several ranking methods, using various heuristic and machine-learning approaches. RESULTS: Our evaluation shows that the annotated document-reaction pairs can be used to create a rule-based document ranking system, and that machine learning can be used to rank documents by their relevance to pathway reactions. We find that a Support Vector Machine-based system outperforms several baselines and matches the performance of the rule-based system. The success of the query extraction and ranking methods are used to update our existing pathway search system, PathText. AVAILABILITY: An online demonstration of PathText 2 and the annotated corpus are available for research purposes at http://www.nactem.ac.uk/pathtext2/. SUPPLEMENTARY INFORMATION: Supplementary data are available at Bioinformatics online.
Asunto(s)
Fenómenos Bioquímicos , Minería de Datos/métodos , Algoritmos , Inteligencia Artificial , MEDLINE , Máquina de Vectores de SoporteRESUMEN
BACKGROUND: Biomedical events are key to understanding physiological processes and disease, and wide coverage extraction is required for comprehensive automatic analysis of statements describing biomedical systems in the literature. In turn, the training and evaluation of extraction methods requires manually annotated corpora. However, as manual annotation is time-consuming and expensive, any single event-annotated corpus can only cover a limited number of semantic types. Although combined use of several such corpora could potentially allow an extraction system to achieve broad semantic coverage, there has been little research into learning from multiple corpora with partially overlapping semantic annotation scopes. RESULTS: We propose a method for learning from multiple corpora with partial semantic annotation overlap, and implement this method to improve our existing event extraction system, EventMine. An evaluation using seven event annotated corpora, including 65 event types in total, shows that learning from overlapping corpora can produce a single, corpus-independent, wide coverage extraction system that outperforms systems trained on single corpora and exceeds previously reported results on two established event extraction tasks from the BioNLP Shared Task 2011. CONCLUSIONS: The proposed method allows the training of a wide-coverage, state-of-the-art event extraction system from multiple corpora with partial semantic annotation overlap. The resulting single model makes broad-coverage extraction straightforward in practice by removing the need to either select a subset of compatible corpora or semantic types, or to merge results from several models trained on different individual corpora. Multi-corpus learning also allows annotation efforts to focus on covering additional semantic types, rather than aiming for exhaustive coverage in any single annotation effort, or extending the coverage of semantic types annotated in existing corpora.