CollaboNet: collaboration of deep neural networks for biomedical named entity recognition

Yoon, Wonjin,So, Chan Ho,Lee, Jinhyuk,Kang, Jaewoo BioMed Central 2019 BMC bioinformatics Vol.20 No.10

<P><B>Background</B></P><P>Finding biomedical named entities is one of the most essential tasks in biomedical text mining. Recently, deep learning-based approaches have been applied to biomedical named entity recognition (BioNER) and showed promising results. However, as deep learning approaches need an abundant amount of training data, a lack of data can hinder performance. BioNER datasets are scarce resources and each dataset covers only a small subset of entity types. Furthermore, many bio entities are polysemous, which is one of the major obstacles in named entity recognition.</P><P><B>Results</B></P><P>To address the lack of data and the entity type misclassification problem, we propose CollaboNet which utilizes a combination of multiple NER models. In CollaboNet, models trained on a different dataset are connected to each other so that a target model obtains information from other collaborator models to reduce false positives. Every model is an expert on their target entity type and takes turns serving as a target and a collaborator model during training time. The experimental results show that CollaboNet can be used to greatly reduce the number of false positives and misclassified entities including polysemous words. CollaboNet achieved state-of-the-art performance in terms of precision, recall and F1 score.</P><P><B>Conclusions</B></P><P>We demonstrated the benefits of combining multiple models for BioNER. Our model has successfully reduced the number of misclassified entities and improved the performance by leveraging multiple datasets annotated for different entity types. Given the state-of-the-art performance of our model, we believe that CollaboNet can improve the accuracy of downstream biomedical text mining applications such as bio-entity relation extraction.</P>

Recent advances in biomedical applications of accelerator mass spectrometry

BioMed Central 2009 Journal of biomedical science Vol.16 No.1

<P>The use of radioisotopes has a long history in biomedical science, and the technique of accelerator mass spectrometry (AMS), an extremely sensitive nuclear physics technique for detection of very low-abundant, stable and long-lived isotopes, has now revolutionized high-sensitivity isotope detection in biomedical research, because it allows the direct determination of the amount of isotope in a sample rather than measuring its decay, and thus the quantitative analysis of the fate of the radiolabeled probes under the given conditions. Since AMS was first used in the early 90's for the analysis of biological samples containing enriched <SUP>14</SUP>C for toxicology and cancer research, the biomedical applications of AMS to date range from <I>in vitro </I>to <I>in vivo </I>studies, including the studies of 1) toxicant and drug metabolism, 2) neuroscience, 3) pharmacokinetics, and 4) nutrition and metabolism of endogenous molecules such as vitamins. In addition, a new drug development concept that relies on the ultrasensitivity of AMS, known as human microdosing, is being used to obtain early human metabolism information of candidate drugs. These various aspects of AMS are reviewed and a perspective on future applications of AMS to biomedical research is provided.</P>

An application of convolutional neural networks with salient features for relation classification

Dashdorj, Zolzaya,Song, Min BioMed Central 2019 BMC bioinformatics Vol.20 No.10

<P><B>Background</B></P><P>Due to the advent of deep learning, the increasing number of studies in the biomedical domain has attracted much interest in feature extraction and classification tasks. In this research, we seek the best combination of feature set and hyperparameter setting of deep learning algorithms for relation classification. To this end, we incorporate an entity and relation extraction tool, PKDE4J to extract biomedical features (i.e., biomedical entities, relations) for the relation classification. We compared the chosen Convolutional Neural Networks (CNN) based classification model with the most widely used learning algorithms.</P><P><B>Results</B></P><P>Our CNN based classification model outperforms the most widely used supervised algorithms. We achieved a significant performance on binary classification with a weighted macro-average F1-score: 94.79% using pre-extracted relevant feature combinations. For multi-class classification, the weighted macro-average F1-score is estimated around 86.95%.</P><P><B>Conclusions</B></P><P>Our results suggest that our proposed CNN based model using the not only single feature as the raw text of the sentences of biomedical literature, but also coupling with multiple and highlighted features extracted from the biomedical sentences could improve the classification performance significantly. We offer hyperparameter tuning and optimization approaches for our proposed model to obtain optimal hyperparameters of the models with the best performance.</P>

Cystatin M loss is associated with the losses of estrogen receptor, progesterone receptor, and HER4 in invasive breast cancer

Ko, Eunkyung,Park, Seong-Eun,Cho, Eun Yoon,Kim, Yujin,Hwang, Jung-Ah,Lee, Yeon-Su,Nam, Seok Jin,Bang, Saik,Park, Joobae,Kim, Duk-Hwan BioMed Central 2010 Breast cancer research Vol.12 No.6

<P><B>Introduction</B></P><P>This study was aimed at understanding the clinicopathological significance of cystatin M loss, and investigating possible factors responsible for cystatin M loss in breast cancer.</P><P><B>Methods</B></P><P>The expression of estrogen receptor (ER), progesterone receptor (PR), HER2, HER4, and cystatin M was retrospectively analyzed using immunohistochemistry in 117 patients with ductal carcinoma <I>in situ </I>(DCIS) and in 175 patients with invasive breast cancer (IBC). The methylation status of <I>CST6 </I>gene encoding cystatin M was evaluated using methylation-specific polymerase chain reaction (PCR) in formalin-fixed paraffin-embedded tissues from 292 participants and using pyrosequencing in fresh-frozen tumor and matched normal tissues from 51 IBC patients.</P><P><B>Results</B></P><P>Cystatin M loss was found in 9 (8%) of 117 patients with DCIS and in 99 (57%) of 175 with invasive breast cancer (IBC) (<I>P </I>< 0.0001). Cystatin M loss was found in 58 (57%) of 101 HER2-negative IBCs and in 41 (55%) of 74 HER2-positive IBCs, and this difference was not statistically significant (<I>P </I>= 0.97). However, cystatin M loss was significantly associated with the loss of ER (<I>P </I>= 0.01), PR (<I>P </I>= 0.002), and HER4 (<I>P </I>= 0.003) in IBCs. Cystatin M loss occurred in 34 (76%) of the 45 HER4-negative IBCs and in 65 (50%) of the 130 HER4-positive IBCs. Multivariate analysis showed that cystatin M loss occurred at a 3.57 times (95% CI = 1.28 to 9.98; <I>P </I>= 0.01) higher prevalence in the triple-negative IBCs of ER, PR, and HER4 than in other subtypes, after adjusting for age. The quantity of <I>CST6 </I>methylation was associated with ER loss (<I>P </I>= 0.0002) in IBCs but not with the loss of PR (<I>P </I>= 0.64) or HER4 (<I>P </I>= 0.87).</P><P><B>Conclusions</B></P><P>The present study suggests that cystatin M loss may be associated with the losses of ER, PR, and HER4 in IBC.</P>

Analysis of AML genes in dysregulated molecular networks

Lee, Eunjung,Jung, Hyunchul,Radivojac, Predrag,Kim, Jong-Won,Lee, Doheon BioMed Central 2009 BMC bioinformatics Vol.10 No.suppl9

<P><B>Background</B></P><P>Identifying disease causing genes and understanding their molecular mechanisms are essential to developing effective therapeutics. Thus, several computational methods have been proposed to prioritize candidate disease genes by integrating different data types, including sequence information, biomedical literature, and pathway information. Recently, molecular interaction networks have been incorporated to predict disease genes, but most of those methods do not utilize invaluable disease-specific information available in mRNA expression profiles of patient samples.</P><P><B>Results</B></P><P>Through the integration of protein-protein interaction networks and gene expression profiles of acute myeloid leukemia (AML) patients, we identified subnetworks of interacting proteins dysregulated in AML and characterized known mutation genes causally implicated to AML embedded in the subnetworks. The analysis shows that the set of extracted subnetworks is a reservoir rich in AML genes reflecting key leukemogenic processes such as myeloid differentiation.</P><P><B>Conclusion</B></P><P>We showed that the integrative approach both utilizing gene expression profiles and molecular networks could identify AML causing genes most of which were not detectable with gene expression analysis alone due to the minor changes in mRNA level.</P>

Carbonic anhydrase XII expression is associated with histologic grade of cervical cancer and superior radiotherapy outcome

Yoo, Chong Woo,Nam, Byung-Ho,Kim, Joo-Young,Shin, Hye-Jin,Lim, Hyunsun,Lee, Sun,Lee, Su-Kyoung,Lim, Myong-Cheol,Song, Yong-Jung BioMed Central 2010 Radiation oncology Vol.5 No.-

<P><B>Background</B></P><P>To investigate whether expression of carbonic anhydrase XII (CA12) is associated with histologic grade of the tumors and radiotherapy outcomes of the patients with invasive cervical cancer.</P><P><B>Methods</B></P><P>CA12 expression was examined by immunohistochemical stains in cervical cancer tissues from 183 radiotherapy patients. Histological grading was classified as well (WD), moderately (MD) or poorly differentiated (PD). Oligonucleotide microarray experiment was performed using seven cervical cancer samples to examine differentially expressed genes between WD and PD cervical cancers. The association between CA12 and histological grade was analyzed by chi-square test. CA12 and histological grades were analyzed individually and as combined CA12 and histologic grade categories for effects on survival outcome.</P><P><B>Results</B></P><P>Immunohistochemical expression of CA12 was highly associated with the histologic grade of cervical cancer. Lack of CA12 expression was associated with PD histology, with an odds ratio of 3.9 (<I>P </I>= 0.01). Microarray analysis showed a fourfold reduction in <I>CA12 </I>gene expression in PD tumors. CA12 expression was marginally associated with superior disease-free survival. Application of the new combined categories resulted in further discrimination of the prognosis of patients with moderate and poorly differentiated tumor grade.</P><P><B>Conclusions</B></P><P>Our study indicates that CA12 may be used as a novel prognostic marker in combination with histologic grade of the tumors.</P>

SynechoNET: integrated protein-protein interaction database of a model cyanobacterium <i>Synechocystis </i> sp. PCC 6803

Kim, Woo-Yeon,Kang, Sungsoo,Kim, Byoung-Chul,Oh, Jeehyun,Cho, Seongwoong,Bhak, Jong,Choi, Jong-Soon BioMed Central 2008 BMC bioinformatics Vol.9 No.suppl1

<P><B>Background</B></P><P>Cyanobacteria are model organisms for studying photosynthesis, carbon and nitrogen assimilation, evolution of plant plastids, and adaptability to environmental stresses. Despite many studies on cyanobacteria, there is no web-based database of their regulatory and signaling protein-protein interaction networks to date.</P><P><B>Description</B></P><P>We report a database and website SynechoNET that provides predicted protein-protein interactions. SynechoNET shows cyanobacterial domain-domain interactions as well as their protein-level interactions using the model cyanobacterium, <I>Synechocystis </I>sp. PCC 6803. It predicts the protein-protein interactions using public interaction databases that contain mutually complementary and redundant data. Furthermore, SynechoNET provides information on transmembrane topology, signal peptide, and domain structure in order to support the analysis of regulatory membrane proteins. Such biological information can be queried and visualized in user-friendly web interfaces that include the interactive network viewer and search pages by keyword and functional category.</P><P><B>Conclusion</B></P><P>SynechoNET is an integrated protein-protein interaction database designed to analyze regulatory membrane proteins in cyanobacteria. It provides a platform for biologists to extend the genomic data of cyanobacteria by predicting interaction partners, membrane association, and membrane topology of <I>Synechocystis </I>proteins. SynechoNET is freely available at or directly at .</P>

Identification of temporal association rules from time-series microarray data sets

Nam, Hojung,Lee, KiYoung,Lee, Doheon BioMed Central 2009 BMC bioinformatics Vol.10 No.suppl3

<P><B>Background</B></P><P>One of the most challenging problems in mining gene expression data is to identify how the expression of any particular gene affects the expression of other genes. To elucidate the relationships between genes, an association rule mining (ARM) method has been applied to microarray gene expression data. However, a conventional ARM method has a limit on extracting temporal dependencies between gene expressions, though the temporal information is indispensable to discover underlying regulation mechanisms in biological pathways. In this paper, we propose a novel method, referred to as temporal association rule mining (TARM), which can extract temporal dependencies among related genes. A temporal association rule has the form [<I>gene A</I>↑, <I>gene B</I>↓] → (7 min) [<I>gene C</I>↑], which represents that high expression level of <I>gene A </I>and significant repression of <I>gene B </I>followed by significant expression of <I>gene C </I>after 7 minutes. The proposed TARM method is tested with <I>Saccharomyces cerevisiae </I>cell cycle time-series microarray gene expression data set.</P><P><B>Results</B></P><P>In the parameter fitting phase of TARM, the fitted parameter set [threshold = ± 0.8, support ≥ 3 transactions, confidence ≥ 90%] with the best precision score for KEGG cell cycle pathway has been chosen for rule mining phase. With the fitted parameter set, numbers of temporal association rules with five transcriptional time delays (0, 7, 14, 21, 28 minutes) are extracted from gene expression data of 799 genes, which are pre-identified cell cycle relevant genes. From the extracted temporal association rules, associated genes, which play same role of biological processes within short transcriptional time delay and some temporal dependencies between genes with specific biological processes are identified.</P><P><B>Conclusion</B></P><P>In this work, we proposed TARM, which is an applied form of conventional ARM. TARM showed higher precision score than Dynamic Bayesian network and Bayesian network. Advantages of TARM are that it tells us the size of transcriptional time delay between associated genes, activation and inhibition relationship between genes, and sets of co-regulators.</P>

Dysregulated balance of Th17 and Th1 cells in systemic lupus erythematosus

Shah, Kamini,Lee, Won-Woo,Lee, Seung-Hyun,Kim, Sang Hyun,Kang, Seong Wook,Craft, Joe,Kang, Insoo BioMed Central 2010 ARTHRITIS RESEARCH AND THERAPY Vol.12 No.2

<P><B>Introduction</B></P><P>Interleukin (IL)-17 is a proinflammatory cytokine that is produced largely by a unique CD4<SUP>+ </SUP>T-helper (Th) subset called Th17 cells. The development of Th17 cells is suppressed by interferon (IFN)-γ produced by Th1 cells, suggesting cross-regulation between Th17 and Th1 cells. Thus, this study analyzed the balance of CD4<SUP>+ </SUP>Th17 and Th1 cell responses in peripheral blood from patients with systemic lupus erythematosus (SLE) and healthy subjects.</P><P><B>Methods</B></P><P>Twenty-five adult patients with SLE and 26 healthy subjects matched for gender and age (± 2 years) were recruited. Peripheral blood mononuclear cells (PBMCs) from patients and healthy subjects were stimulated for 4 h <I>ex vivo </I>with phorbol myristate acetate (PMA) and ionomycin. The frequency of CD4<SUP>+ </SUP>T cells producing IL-17 and/or IFN-γ was measured by using flow cytometry. Expression of Th17-associated chemokine receptors CCR4 and CCR6 on CD4<SUP>+ </SUP>T cells as well as plasma levels of Th17-polarizing cytokines were assessed. Disease activity was evaluated by the SLE disease activity index score (SLEDAI). Unpaired <I>t </I>test and Pearson correlation were used for statistical analyses.</P><P><B>Results</B></P><P>Patients with SLE had an increased frequency of CD4<SUP>+</SUP>IL-17<SUP>+ </SUP>T cells compared with healthy subjects. However, the frequency of CD4<SUP>+</SUP>IFN-γ<SUP>+ </SUP>T cells was similar between the two groups, indicating an altered balance of Th17 and Th1 cell responses in SLE. Patients with SLE also had an increased frequency of CD4<SUP>+</SUP>CCR4<SUP>+</SUP>CCR6<SUP>+ </SUP>T cells that are known to produce IL-17. The frequency of CD4<SUP>+</SUP>IL-17<SUP>+ </SUP>T cells and CD4<SUP>+</SUP>CCR4<SUP>+</SUP>CCR6<SUP>+ </SUP>T cells correlated with disease activity. In measuring plasma levels of the Th17-polarizing cytokines, levels of IL-6 were higher in patients with SLE than in healthy subjects, although levels of IL-1β, IL-21, IL-23, and transforming growth factor (TGF)-β were not different between the two groups.</P><P><B>Conclusions</B></P><P>We demonstrate an enhanced Th17 cell response that correlates with disease activity in patients with SLE, suggesting a role for IL-17 in the pathogenesis of lupus. Our data indicate that the mechanisms involved in balancing Th1 and Th17 regulation, as well as in producing IL-6, are aberrant in SLE, leading to an increased Th17 response. We suggest that CCR4 and CCR6 expression on CD4<SUP>+ </SUP>T cells should be considered as markers of disease activity, and that IL-17 blocking may offer a therapeutic target in SLE.</P>

Cleavage of ST6Gal I by Radiation-Induced BACE1 Inhibits Golgi-Anchored ST6Gal I-Mediated Sialylation of Integrin β1 and Migration in Colon Cancer Cells

Lee, Minyoung,Park, Jung-Jin,Ko, Young-Gyu,Lee, Yun-Sil BioMed Central 2012 Radiation oncology Vol.7 No.-

<P><B>Background</B></P><P>Previously, we found that β-galactoside α2,6-sialyltransferase (ST6Gal I), an enzyme that adds sialic acids to N-linked oligosaccharides of glycoproteins and is frequently overexpressed in cancer cells, is up-regulated by ionizing radiation (IR) and cleaved to a form possessing catalytic activity comparable to that of the Golgi-localized enzyme. Moreover, this soluble form is secreted into the culture media. Induction of ST6Gal I significantly increased the migration of colon cancer cells via sialylation of integrin β1. Here, we further investigated the mechanisms underlying ST6Gal I cleavage, solubilization and release from cells, and addressed its functions, focusing primarily on cancer cell migration.</P><P><B>Methods</B></P><P>We performed immunoblotting and lectin affinity assay to analyze the expression of ST6 Gal I and level of sialylated integrin β1. After ionizing radiation, migration of cells was measured by in vitro migration assay. α2, 6 sialylation level of cell surface was analyzed by flow cytometry. Cell culture media were concentrated and then analyzed for soluble ST6Gal I levels using an α2, 6 sialyltransferase sandwich ELISA.</P><P><B>Result</B></P><P>We found that ST6Gal I was cleaved by BACE1 (β-site amyloid precursor protein-cleaving enzyme), which was specifically overexpressed in response to IR. The soluble form of ST6Gal I, which also has sialyltransferase enzymatic activity, was cleaved from the Golgi membrane and then released into the culture media. Both non-cleaved and cleaved forms of ST6Gal I significantly increased colon cancer cell migration in a sialylation-dependent manner. The pro-migratory effect of the non-cleaved form of ST6Gal I was dependent on integrin β1 sialylation, whereas that of the cleaved form of ST6Gal I was not, suggesting that other intracellular sialylated molecules apart from cell surface molecules such as integrin β1 might be involved in mediating the pro-migratory effects of the soluble form of ST6Gal I. Moreover, production of soluble form ST6Gal I by BACE 1 inhibited integrin β1 sialylation and migration by Golgi-anchored form of ST6Gal I.</P><P><B>Conclusions</B></P><P>Our results suggest that soluble ST6Gal I, possibly in cooperation with the Golgi-bound form, may participate in cancer progression and metastasis prior to being secreted from cancer cells.</P>