Valproic acid monitoring: Serum prediction using a machine learning framework from multicenter real-world data

BACKGROUND

Our study employs machine learning to predict serum valproic acid (VPA) concentrations, aiming to contribute to the development of non-invasive assays for therapeutic drug monitoring.

METHODS

Medical records from 2002 to 2019 were obtained from the Taiwan Chang Gung Research Database. Using various machine learning algorithms, we developed predictive models to classify serum VPA concentrations into two categories (1-50 μg/ml or 51-100 μg/ml) and predicted the exact concentration value. The models were trained on 5142 samples and tested on 644 independent samples. Accuracy was the main metric used to evaluate model performance, with a tolerance of 20 μg/ml for continuous variables. Furthermore, we identified important features and developed simplified models with fewer features.

RESULTS

The models achieved an average accuracy of 0.80-0.86 for binary outcomes and 0.72-0.88 for continuous outcome. Ten top features associated with higher serum VPA levels included higher VPA last and daily doses, bipolar disorder or schizophrenia spectrum disorder diagnoses, elevated levels of serum albumin, calcium, and creatinine, low platelet count, low percentage of segmented white blood cells, and low red cell distribution width-coefficient of variation. The simplified models had an average accuracy of 0.82-0.86 for binary outcome and 0.70-0.86 for continuous outcome.

LIMITATIONS

The study's predictive model lacked external test data from outside the hospital for validation.

CONCLUSIONS

Machine learning models have the potential to integrate real-world data and predict VPA concentrations, providing a promising tool for reducing the need for frequent monitoring of serum levels in clinical practice.

Predicting new-onset post-stroke depression from real-world data using machine learning algorithm

Introduction

Post-stroke depression (PSD) is a serious mental disorder after ischemic stroke. Early detection is important for clinical practice. This research aims to develop machine learning models to predict new-onset PSD using real-world data.

Methods

We collected data for ischemic stroke patients from multiple medical institutions in Taiwan between 2001 and 2019. We developed models from 61,460 patients and used 15,366 independent patients to test the models' performance by evaluating their specificities and sensitivities. The predicted targets were whether PSD occurred at 30, 90, 180, and 365 days post-stroke. We ranked the important clinical features in these models.

Results

In the study's database sample, 1.3% of patients were diagnosed with PSD. The average specificity and sensitivity of these four models were 0.83-0.91 and 0.30-0.48, respectively. Ten features were listed as important features related to PSD at different time points, namely old age, high height, low weight post-stroke, higher diastolic blood pressure after stroke, no pre-stroke hypertension but post-stroke hypertension (new-onset hypertension), post-stroke sleep-wake disorders, post-stroke anxiety disorders, post-stroke hemiplegia, and lower blood urea nitrogen during stroke.

Discussion

Machine learning models can provide as potential predictive tools for PSD and important factors are identified to alert clinicians for early detection of depression in high-risk stroke patients.

LAD: Layer-Wise Adaptive Distillation for BERT Model Compression

Recent advances with large-scale pre-trained language models (e.g., BERT) have brought significant potential to natural language processing. However, the large model size hinders their use in IoT and edge devices. Several studies have utilized task-specific knowledge distillation to compress the pre-trained language models. However, to reduce the number of layers in a large model, a sound strategy for distilling knowledge to a student model with fewer layers than the teacher model is lacking. In this work, we present Layer-wise Adaptive Distillation (LAD), a task-specific distillation framework that can be used to reduce the model size of BERT. We design an iterative aggregation mechanism with multiple gate blocks in LAD to adaptively distill layer-wise internal knowledge from the teacher model to the student model. The proposed method enables an effective knowledge transfer process for a student model, without skipping any teacher layers. The experimental results show that both the six-layer and four-layer LAD student models outperform previous task-specific distillation approaches during GLUE tasks.

HAT4RD: Hierarchical Adversarial Training for Rumor Detection in Social Media

With the development of social media, social communication has changed. While this facilitates people's communication and access to information, it also provides an ideal platform for spreading rumors. In normal or critical situations, rumors can affect people's judgment and even endanger social security. However, natural language is high-dimensional and sparse, and the same rumor may be expressed in hundreds of ways on social media. As such, the robustness and generalization of the current rumor detection model are in question. We proposed a novel hierarchical adversarial training method for rumor detection (HAT4RD) on social media. Specifically, HAT4RD is based on gradient ascent by adding adversarial perturbations to the embedding layers of post-level and event-level modules to deceive the detector. At the same time, the detector uses stochastic gradient descent to minimize the adversarial risk to learn a more robust model. In this way, the post-level and event-level sample spaces are enhanced, and we verified the robustness of our model under a variety of adversarial attacks. Moreover, visual experiments indicate that the proposed model drifts into an area with a flat loss landscape, thereby, leading to better generalization. We evaluate our proposed method on three public rumor datasets from two commonly used social platforms (Twitter and Weibo). Our experimental results demonstrate that our model achieved better results compared with the state-of-the-art methods.

Comparative effectiveness of antidepressants on geriatric depression: Real-world evidence from a population-based study

BACKGROUND

There is little real-world evidence about effectiveness of different antidepressants on geriatric depression.

METHODS

We used population-based claims data in Taiwan between 1997 and 2013 to include older patients (≥ 60 years of age) who were diagnosed with depression and started to use antidepressants. All patients were followed up until discontinuation of antidepressant use or the end of the study period. Treatment outcomes were set as the risk of switching to another antidepressant, receiving augmentation therapy, and psychiatric hospitalization. We used cox proportional hazards regression models to calculate hazard ratios with 95% confidence intervals (CIs) and adjust for several confounding factors (aHRs).

RESULTS

During the study period, a total of 207,946 elderly patients with depression received one of the following 11 antidepressants: sertraline, fluoxetine, paroxetine, escitalopram, citalopram, fluvoxamine, venlafaxine, duloxetine, moclobemide, mirtazapine, and bupropion. Compared to the patients treated with sertraline, those treated with fluvoxamine / venlafaxine had significantly but modestly higher risks of switching (aHR [95% CI]: 1.16 [1.11-1.21] / 1.10 [1.06-1.14]), augmentation (1.06 [1.02-1.10] / 1.08 [1.05-1.12]), and hospitalization (1.28 [1.03-1.58] / 1.37 [1.16-1.62]). Otherwise, the remaining 9 antidepressants yielded no consistent result in the three outcomes.

LIMITATIONS

This study is a multi-arm and active controlled trial, lacking a placebo group.

CONCLUSION

As treating geriatric depression, no individual antidepressant posed consistently better effectiveness in the outcomes of switching antidepressant, receiving augmentation, and psychiatric hospitalization than any other one, whereas clinicians should be cautious when prescribing fluvoxamine and venlafaxine.

Month of birth and the risk of narcolepsy: a systematic review and meta-analysis

STUDY OBJECTIVES

The aim of this study is to evaluate the relationship between the month of birth (MOB) and the risk of narcolepsy.

METHODS

We conducted a systematic review of electronic databases, namely PubMed, Embase, and Cochrane CENTRAL, from their inception to September 30, 2021. We also added data on narcolepsy from the National Database in Taiwan. Then we extracted the relative risk ratios (RR) of narcolepsy in each month of birth to that of the general population and transformed them from month of birth to season. A random-effects model was used to calculate pooled RRs from the meta-analysis and 95% confidence interval (CI).

RESULTS

The current meta-analysis analyzed seven studies and included 3776 patients from eight areas. The RR was highest in March (RR 1.11 [95% CI 0.99-1.26]) or August (1.11 [0.98-1.26]) and lowest in April (0.90 [0.78-1.03]). However, none of the MOBs reached statistical significance. Moreover, the patterns of the three continents were different. In North America, the highest and lowest significant risks were found in March (1.47 [1.20-1.79]) and September (0.75 [95% CI 0.56-0.99]). In Asia, the notable lowest risk was in April (0.80 [0.66-0.97]). In Europe, the risk of narcolepsy is not significantly related to any MOB. In terms of seasons, only spring births in North America had a significantly higher risk (1.21 [1.06-1.38]).

CONCLUSIONS

The findings indicated that the risk of narcolepsy and MOB differed across the three continents. This study indicates the important role of environmental factors in narcolepsy.

SYSTEMATIC REVIEW REGISTRATION

Registry: PROSPERO; Identifier: CRD42020186660.

Predicting Serum Levels of Lithium-Treated Patients: A Supervised Machine Learning Approach

Routine monitoring of lithium levels is common clinical practice. This is because the lithium prediction strategies available developed by previous studies are still limited due to insufficient prediction performance. Thus, we used machine learning approaches to predict lithium concentration in a large real-world dataset. Real-world data from multicenter electronic medical records were used in different machine learning algorithms to predict: (1) whether the serum level was 0.6–1.2 mmol/L or 0.0–0.6 mmol/L (binary prediction), and (2) its concentration value (continuous prediction). We developed models from 1505 samples through 5-fold cross-validation and used 204 independent samples to test their performance by evaluating their accuracy. Moreover, we ranked the most important clinical features in different models and reconstructed three reduced models with fewer clinical features. For binary and continuous predictions, the average accuracy of these models was 0.70–0.73 and 0.68–0.75, respectively. Seven features were listed as important features related to serum lithium levels of 0.6–1.2 mmol/L or higher lithium concentration, namely older age, lower systolic blood pressure, higher daily and last doses of lithium prescription, concomitant psychotropic drugs with valproic acid and -pine drugs, and comorbid substance-related disorders. After reducing the features in the three new predictive models, the binary or continuous models still had an average accuracy of 0.67–0.74. Machine learning processes complex clinical data and provides a potential tool for predicting lithium concentration. This may help in clinical decision-making and reduce the frequency of serum level monitoring.

Differences in Psychiatric Comorbidities and Gender Distribution among Three Clusters of Personality Disorders: A Nationwide Population-Based Study

Personality disorders (PDs) are grouped into clusters A, B, and C. However, whether the three clusters of PDs have differences in comorbid mental disorders or gender distribution is still lacking sufficient evidence. We aim to investigate the distribution pattern across the three clusters of PDs with a population-based cohort study. This study used the Taiwan national database between 1995 and 2013 to examine the data of patients with cluster A PDs, cluster B PDs, or cluster C PDs. We compared the differences of psychiatric comorbidities classified in the Diagnostic and Statistical Manual of Mental Disorders, fifth edition across the three clusters of PDs. Moreover, we formed gender subgroups of the three PDs to observe the discrepancy between male and female. Among the 9845 patients, those with cluster A PDs had the highest proportion of neurodevelopmental disorders, schizophrenia and neurocognitive disorders, those with cluster B PDs demonstrated the largest percentage of bipolar disorders, trauma and stressor disorders, feeding and eating disorders, and substance and addictive disorders, and those with cluster C PDs had the greatest proportion of depressive disorders, anxiety disorders, obsessive–compulsive disorders, somatic symptom disorders, and sleep–wake disorders. The gender subgroups revealed significant male predominance in neurodevelopmental disorders and female predominance in sleep–wake disorders across all three clusters of PDs. Our findings support that some psychiatric comorbidities are more prevalent in specified cluster PDs and that gender differences exist across the three clusters of PDs. These results are an important reference for clinicians who are developing services that target real-world patients with PDs.

An Innovative Wearable Device For Monitoring Continuous Body Surface Temperature (HEARThermo): Instrument Validation Study

Background

Variations in body temperature are highly informative during an illness. To date, there are not many adequate studies that have investigated the feasibility of a wearable wrist device for the continuous monitoring of body surface temperatures in humans.

Objective

The objective of this study was to validate the performance of HEARThermo, an innovative wearable device, which was developed to continuously monitor the body surface temperature in humans.

Methods

We implemented a multi-method research design in this study, which included 2 validation studies—one in the laboratory and one with human subjects. In validation study I, we evaluated the test-retest reliability of HEARThermo in the laboratory to measure the temperature and to correct the values recorded by each HEARThermo by using linear regression models. We conducted validation study II on human subjects who wore HEARThermo for the measurement of their body surface temperatures. Additionally, we compared the HEARThermo temperature recordings with those recorded by the infrared skin thermometer simultaneously. We used intraclass correlation coefficients (ICCs) and Bland-Altman plots to analyze the criterion validity and agreement between the 2 measurement tools.

Results

A total of 66 participants (age range, 10-77 years) were recruited, and 152,881 completed data were analyzed in this study. The 2 validation studies in the laboratory and on human skin indicated that HEARThermo showed a good test-retest reliability (ICC 0.96-0.98) and adequate criterion validity with the infrared skin thermometer at room temperatures of 20°C-27.9°C (ICC 0.72, P<.001). The corrected measurement bias averaged –0.02°C, which was calibrated using a water bath ranging in temperature from 16°C to 40°C. The values of each HEARThermo improved by the regression models were not significantly different from the temperature of the water bath (P=.19). Bland-Altman plots showed no visualized systematic bias. HEARThermo had a bias of 1.51°C with a 95% limit of agreement between –1.34°C and 4.35°C.

Conclusions

The findings of our study show the validation of HEARThermo for the continuous monitoring of body surface temperatures in humans.

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.

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.

Genealogical-based method for multiple ontology self-extension in MeSH

During the last decade, the advent of Ontologies used for biomedical annotation has had a deep impact on life science. MeSH is a well-known Ontology for the purpose of indexing journal articles in PubMed, improving literature searching on multi-domain topics. Since the explosion of data growth in recent years, there are new terms, concepts that weed through the old and bring forth the new. Automatically extending sets of existing terms will enable bio-curators to systematically improve text-based ontologies level by level. However, most of the related techniques which apply symbolic patterns based on a literature corpus tend to focus on more general but not specific parts of the ontology. Therefore, in this work, we present a novel method for utilizing genealogical information from Ontology itself to find suitable siblings for ontology extension. Based on the breadth and depth dimensions, the sibling generation stage and pruning strategy are proposed in our approach. As a result, on the average, the precision of the genealogical-based method achieved 0.5, with the best 0.83 performance of category "Organisms." We also achieve average precision 0.69 of 229 new terms in MeSH 2013 version.

Unsupervised corpus distillation for represented indicator measurement on focus species detection

The gene ambiguity with the highest dimension is the species with which an entity is associated in biomedical text mining. Furthermore, one of the bottlenecks in gene normalisation is focus species detection. This study presents a method which is robust for all types of articles, particularly those without explicit species mentions. Since our method requires a training corpus, we developed an iterative distillation method to extend the corpus. Unsupervised corpus is therefore helpful for the detection of focus species. In experiments, the proposed method achieved a high accuracy of 85.64% and 84.32% in datasets with and without species mentions respectively.

A semi-supervised, weighted pattern-learning approach for extraction of gene regulation relationships from scientific literature

Moreover, the large amount of textual knowledge in the existing biomedical literature is growing rapidly, and the creation of manual patterns from the available literature is becoming more difficult. There is an increasing demand to extract potential generic regulatory relationships from unlabelled data sets. In this paper, we describe a Semi-Supervised, Weighted Pattern Learning method (SSWPL) to extract such generic regulatory information from the literature. SSWPL can build new regulatory patterns according to predefined initial patterns from unlabelled data in the literature. These constructed regulatory patterns are then used to extract generic regulatory information from PubMed abstracts. The results presented herein demonstrate that our method can be utilised to effectively extract generic regulatory relationships from the literature by using learned, weighted patterns through semi-supervised pattern learning.

Gene Name Disambiguation Using Multi-Scope Species Detection

Species detection is an important topic in the text mining field. According to the importance of the research topics (e.g., species assignment to genes and document focus species detection), some studies are dedicated to an individual topic. However, no researcher to date has discussed species detection as a general problem. Therefore, we developed a multi-scope species detection model to identify the focus species for different scopes (i.e., gene mention, sentence, paragraph, and global scope of the entire article). Species assignment is one of the bottlenecks of gene name disambiguation. In our evaluation, recognizing the focus species of a gene mention in four different scopes improved the gene name disambiguation. We used the species cue words extracted from articles to estimate the relevance between an article and a species. The relevance score was calculated by our proposed entities frequency-augmented invert species frequency (EF-AISF) formula, which represents the importance of an entity to a species. We also defined a relation guide factor (RGF) to normalize the relevance score. Our method not only achieved better performance than previous methods but also can handle the articles that do not specifically mention a species. In the DECA corpus, we outperformed previous studies and obtained an accuracy of 88.22 percent.

Using a search engine-based mutually reinforcing approach to assess the semantic relatedness of biomedical terms

Background

Determining the semantic relatedness of two biomedical terms is an important task for many text-mining applications in the biomedical field. Previous studies, such as those using ontology-based and corpus-based approaches, measured semantic relatedness by using information from the structure of biomedical literature, but these methods are limited by the small size of training resources. To increase the size of training datasets, the outputs of search engines have been used extensively to analyze the lexical patterns of biomedical terms.

Methodology/Principal Findings

In this work, we propose the Mutually Reinforcing Lexical Pattern Ranking (ReLPR) algorithm for learning and exploring the lexical patterns of synonym pairs in biomedical text. ReLPR employs lexical patterns and their pattern containers to assess the semantic relatedness of biomedical terms. By combining sentence structures and the linking activities between containers and lexical patterns, our algorithm can explore the correlation between two biomedical terms.

Conclusions/Significance

The average correlation coefficient of the ReLPR algorithm was 0.82 for various datasets. The results of the ReLPR algorithm were significantly superior to those of previous methods.

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.

PubTator: a web-based text mining tool for assisting biocuration

Manually curating knowledge from biomedical literature into structured databases is highly expensive and time-consuming, making it difficult to keep pace with the rapid growth of the literature. There is therefore a pressing need to assist biocuration with automated text mining tools. Here, we describe PubTator, a web-based system for assisting biocuration. PubTator is different from the few existing tools by featuring a PubMed-like interface, which many biocurators find familiar, and being equipped with multiple challenge-winning text mining algorithms to ensure the quality of its automatic results. Through a formal evaluation with two external user groups, PubTator was shown to be capable of improving both the efficiency and accuracy of manual curation. PubTator is publicly available at http://www.ncbi.nlm.nih.gov/CBBresearch/Lu/Demo/PubTator/.

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

Accelerating literature curation with text-mining tools: a case study of using PubTator to curate genes in PubMed abstracts

Today’s biomedical research has become heavily dependent on access to the biological knowledge encoded in expert curated biological databases. As the volume of biological literature grows rapidly, it becomes increasingly difficult for biocurators to keep up with the literature because manual curation is an expensive and time-consuming endeavour. Past research has suggested that computer-assisted curation can improve efficiency, but few text-mining systems have been formally evaluated in this regard. Through participation in the interactive text-mining track of the BioCreative 2012 workshop, we developed PubTator, a PubMed-like system that assists with two specific human curation tasks: document triage and bioconcept annotation. On the basis of evaluation results from two external user groups, we find that the accuracy of PubTator-assisted curation is comparable with that of manual curation and that PubTator can significantly increase human curatorial speed. These encouraging findings warrant further investigation with a larger number of publications to be annotated.

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

Augmented transitive relationships with high impact protein distillation in protein interaction prediction

Predicting new protein-protein interactions is important for discovering novel functions of various biological pathways. Predicting these interactions is a crucial and challenging task. Moreover, discovering new protein-protein interactions through biological experiments is still difficult. Therefore, it is increasingly important to discover new protein interactions. Many studies have predicted protein-protein interactions, using biological features such as Gene Ontology (GO) functional annotations and structural domains of two proteins. In this paper, we propose an augmented transitive relationships predictor (ATRP), a new method of predicting potential protein interactions using transitive relationships and annotations of protein interactions. In addition, a distillation of virtual direct protein-protein interactions is proposed to deal with unbalanced distribution of different types of interactions in the existing protein-protein interaction databases. Our results demonstrate that ATRP can effectively predict protein-protein interactions. ATRP achieves an 81% precision, a 74% recall and a 77% F-measure in average rate in the prediction of direct protein-protein interactions. Using the generated benchmark datasets from KUPS to evaluate of all types of the protein-protein interaction, ATRP achieved a 93% precision, a 49% recall and a 64% F-measure in average rate. This article is part of a Special Issue entitled: Computational Methods for Protein Interaction and Structural Prediction.

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.

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.

Mechanism for nucleocytoplasmic shuttling of histone deacetylase 7

Here we show that HDAC7, a member of the class II histone deacetylases, specifically targets several members of myocyte enhancer factors, MEF2A, -2C, and -2D, and inhibits their transcriptional activity. Furthermore, we demonstrate that DNA-bound MEF2C is capable of recruiting HDAC7, demonstrating that the HDAC7-dependent repression of transcription is not due to the inhibition of the MEF2 DNA binding activity. The data also suggest that the promoter bound MEF2 is potentially capable of remodeling adjacent nucleosomes via the recruitment of HDAC7. We have also observed a nucleocytoplasmic shuttling of HDAC7 and dissected the mechanism involved. In NIH3T3 cells, HDAC7 was primarily localized in the cytoplasm, essentially due to an active CRM1-dependent export of the protein from the nucleus. Interestingly, in HeLa cells, HDAC7 was predominantly nuclear. In these cells we could restore the cytoplasmic localization of HDAC7 by expressing CaMK I. This CaMK I-induced nuclear export of HDAC7 was abolished when three critical serines, Ser-178, Ser-344, and Ser-479, of HDAC7 were mutated. We show that these serines are involved in the direct interaction of HDAC7 with 14-3-3. Mutations of these serine residues weakened the association with 14-3-3 and dramatically enhanced the repression activity of HDAC7 in NIH3T3 cells, but not in HeLa cells. Data presented in this work clearly show that the signal dependent subcellular localization of HDAC7 is essential in controlling its activities. The data also show that the cellular concentration of factors such as 14-3-3, CaMK I, and other yet unknown molecules may determine the subcellular localization of an individual HDAC member in a cell type and HDAC-specific manner.

Sharp, an inducible cofactor that integrates nuclear receptor repression and activation

A yeast two-hybrid screen using the conserved carboxyl terminus of the nuclear receptor corepressor SMRT as a bait led to the isolation of a novel human gene termed SHARP (SMRT/HDAC1 Associated Repressor Protein). SHARP is a potent transcriptional repressor whose repression domain (RD) interacts directly with SMRT and at least five members of the NuRD complex including HDAC1 and HDAC2. In addition, SHARP binds to the steroid receptor RNA coactivator SRA via an intrinsic RNA binding domain and suppresses SRA-potentiated steroid receptor transcription activity. Accordingly, SHARP has the capacity to modulate both liganded and nonliganded nuclear receptors. Surprisingly, the expression of SHARP is itself steroid inducible, suggesting a simple feedback mechanism for attenuation of the hormonal response.

Histone deacetylase complexes: functional entities or molecular reservoirs

Recent investigations have allowed the identification of an increasing number of distinct nuclear multi-component complexes containing several types of enzymatic activity. Many of the complexes containing histone deacetylases are believed to control transcription and chromatin remodeling. We suggest here that at least some of these abundant complexes are likely to be "molecular reservoirs" of dynamic composition that exchange factors with other less abundant and functional complexes, according to specific required activities.

Measurement and pharmacokinetic analysis of unbound cephaloridine in rat blood by on-line microdialysis and microbore liquid chromatography

A technique involving rapid sampling of cephaloridine in rat blood was achieved using a combination of microdialysis and sensitive microbore liquid chromatography. A microdialysis probe was inserted into the jugular vein/right atrium of a Sprague-Dawley rat. Then after a real-time collection of the analyte by microdialysis, the dialysate was automatically injected into a liquid chromatographic system via an on-line injector. Following a 2 h stabilization period after the surgical procedure, cephaloridine (20 mg/kg, i.v.) was then administered via the femoral vein. Isocratic elution of cephaloridine was carried out with a mobile phase containing methanol-20 mM monosodium phosphate (25:75, v/v, pH 5.5), and the flow rate of the mobile phase was 0.05 mL/min within 10 min. Intra- and inter-assay accuracy and precision of the assay were each less than 10%. The in vivo recovery of the cephaloridine from the microdialysate was 49.7 +/- 8.0% and 42.4 +/- 8.4% for 0.5 and 1 microg/mL standards (n = 6), respectively. Based on the pharmacokinetic analysis, the elimination half-life was 32.2 +/- 8.6 min by cephaloridine administration (20 mg/kg, i.v., n = 6).

Measurement and pharmacokinetic analysis of unbound ceftazidime in rat blood using microdialysis and microbore liquid chromatography

To evaluate the biodisposition of ceftazidime in rat blood, a rapid and simple microbore liquid chromatographic technique together with a microdialysis sampling technique were developed. This method involves an on-line design for blood dialysate directly injected into a microbore liquid chromatographic system. The chromatographic conditions consisted of a mobile phase of methanol-acetonitrile-100 mM monosodium phosphoric acid (pH 3.0) (10:10:80, v/v/v) pumped through a microbore reversed-phase column at a flow-rate of 0.05 ml/min. With the detection wavelength set at 254 nm, a good linear correlation was observed between the peak area and the ceftazidime concentration at 0.1 to 50 microg/ml (r=0.999). Microdialysis probes, being custom-made, were screened for acceptable in vivo recovery while chromatographic resolution and detection were validated for response linearity, as well as intra-day and inter-day variabilities. This method was then applied to the pharmacokinetic profiling of ceftazidime in blood following intravenous 50 mg/kg administration to rats. The pharmacokinetics was calculated from the corrected data for dialysate concentrations of ceftazidime versus time. This method has been used to study ceftazidime pharmacokinetics in rats and has proven to be rapid and reproducible.

Identification of a nuclear domain with deacetylase activity

Here, we describe the identification and characterization of a nuclear body (matrix-associated deacetylase body) whose formation and integrity depend on deacetylase activity. Typically, there are 20-40 0.5-microM bodies per nucleus, although the size and number can vary substantially. The structure appears to contain both class I and the recently described class II histone deacetylases (HDAC)5 and 7 along with the nuclear receptor corepressors SMRT (silencing mediator for retinoid and thyroid receptor) and N-CoR (nuclear receptor corepressor). Addition of the deacetylase inhibitors trichostatin A and sodium butyrate completely disrupt these nuclear bodies, providing a demonstration that the integrity of a nuclear body is enzyme dependent. We demonstrate that HDAC5 and 7 can associate with at least 12 distinct proteins, including several members of the NuRD and Sin3A repression complexes, and appear to define a new but related complex.

Isolation of a novel histone deacetylase reveals that class I and class II deacetylases promote SMRT-mediated repression

The transcriptional corepressor SMRT functions by mediating the repressive effect of transcription factors involved in diverse signaling pathways. The mechanism by which SMRT represses basal transcription has been proposed to involve the indirect recruitment of histone deacetylase HDAC1 via the adaptor mSin3A. In contrast to this model, a two-hybrid screen on SMRT-interacting proteins resulted in the isolation of the recently described HDAC5 and a new family member termed HDAC7. Molecular and biochemical results indicate that this interaction is direct and in vivo evidence colocalizes SMRT, mHDAC5, and mHDAC7 to a distinct nuclear compartment. Surprisingly, HDAC7 can interact with mSin3A in yeast and in mammalian cells, suggesting association of multiple repression complexes. Taken together, our results provide the first evidence that SMRT-mediated repression is promoted by class I and class II histone deacetylases and that SMRT can recruit class II histone deacetylases in a mSin3A-independent fashion.

Mechanism of corepressor binding and release from nuclear hormone receptors

The association of transcription corepressors SMRT and N-CoR with retinoid and thyroid receptors results in suppression of basal transcriptional activity. A key event in nuclear receptor signaling is the hormone-dependent release of corepressor and the recruitment of coactivator. Biochemical and structural studies have identified a universal motif in coactivator proteins that mediates association with receptor LBDs. We report here the identity of complementary acting signature motifs in SMRT and N-CoR that are sufficient for receptor binding and ligand-induced release. Interestingly, the motif contains a hydrophobic core (PhixxPhiPhi) similar to that found in NR coactivators. Surprisingly, mutations in the amino acids that directly participate in coactivator binding disrupt the corepressor association. These results indicate a direct mechanistic link between activation and repression via competition for a common or at least partially overlapping binding site.

Pbx-Hox heterodimers recruit coactivator-corepressor complexes in an isoform-specific manner

Homeobox (hox) proteins have been shown to regulate cell fate and segment identity by promoting the expression of specific genetic programs. In contrast to their restricted biological action in vivo, however, most homeodomain factors exhibit promiscuous DNA binding properties in vitro, suggesting a requirement for additional cofactors that enhance target site selectivity. In this regard, the pbx family of homeobox genes has been found to heterodimerize with and thereby augment the DNA binding activity of certain hox proteins on a subset of potential target sites. Here we examine the transcriptional properties of a forced hox-pbx heterodimer containing the pancreas-specific orphan homeobox factor pdx fused to pbx-1a. Compared to the pdx monomer, the forced pdx-pbx1a dimer, displayed 10- to 20-fold-higher affinity for a consensus hox-pbx binding site but was completely unable to bind a hox monomer recognition site. The pdx-pbx dimer stimulated target gene expression via an N-terminal trans-activation domain in pdx that interacts with the coactivator CREB binding protein. The pdx-pbx dimer was also found to repress transcription via a C-terminal domain in pbx-1a that associates with the corepressors SMRT and NCoR. The transcriptional properties of the pdx-pbx1 complex appear to be regulated at the level of alternative splicing; a pdx-pbx polypeptide containing the pbx1b isoform, which lacks the C-terminal extension in pbx1a, was unable to repress target gene expression via NCoR-SMRT. Since pbx1a and pbx1b are differentially expressed in endocrine versus exocrine compartments of the adult pancreas, our results illustrate a novel mechanism by which pbx proteins may modulate the expression of specific genetic programs, either positively or negatively, during development.

SMRTER, a Drosophila nuclear receptor coregulator, reveals that EcR-mediated repression is critical for development

The Drosophila ecdysone receptor (EcR)/ultraspiracle (USP) heterodimer is a key regulator in molting and metamorphoric processes, activating and repressing transcription in a sequence-specific manner. Here, we report the isolation of an EcR-interacting protein, SMRTER, which is structurally divergent but functionally similar to the vertebrate nuclear corepressors SMRT and N-CoR. SMRTER mediates repression by interacting with Sin3A, a repressor known to form a complex with the histone deacetylase Rpd3/HDAC. Importantly, we identify an EcR mutant allele that fails to bind SMRTER and is characterized by developmental defects and lethality. Together, these results reveal a novel nuclear receptor cofactor that exhibits evolutionary conservation in the mechanism to achieve repression and demonstrate the essential role of repression in hormone signaling.

A viral mechanism for inhibition of p300 and PCAF acetyltransferase activity

Nucleosomal histone modification is believed to be a critical step in the activation of RNA polymerase II-dependent transcription. p300/CBP and PCAF histone acetyltransferases (HATs) are coactivators for several transcription factors, including nuclear hormone receptors, p53, and Stat1alpha, and participate in transcription by forming an activation complex and by promoting histone acetylation. The adenoviral E1A oncoprotein represses transcriptional signaling by binding to p300/CBP and displacing PCAF and p/CIP proteins from the complex. Here, we show that E1A directly represses the HAT activity of both p300/CBP and PCAF in vitro and p300-dependent transcription in vivo. Additionally, E1A inhibits nucleosomal histone modifications by the PCAF complex and blocks p53 acetylation. These results demonstrate the modulation of HAT activity as a novel mechanism of transcriptional regulation.

A histone deacetylase corepressor complex regulates the Notch signal transduction pathway

The Delta-Notch signal transduction pathway has widespread roles in animal development in which it appears to control cell fate. CBF1/RBP-Jkappa, the mammalian homolog of Drosophila Suppressor of Hairless [Su(H)], switches from a transcriptional repressor to an activator upon Notch activation. The mechanism whereby Notch regulates this switch is not clear. In this report we show that prior to induction CBF1/RBP-Jkappa interacts with a corepressor complex containing SMRT (silencing mediator of retinoid and thyroid hormone receptors) and the histone deacetylase HDAC-1. This complex binds via the CBF1 repression domain, and mutants defective in repression fail to interact with the complex. Activation by Notch disrupts the formation of the repressor complex, thus establishing a molecular basis for the Notch switch. Finally, ESR-1, a Xenopus gene activated by Notch and X-Su(H), is induced in animal caps treated with TSA, an inhibitor of HDAC-1. The functional role for the SMRT/HDAC-1 complex in CBF1/RBP-Jkappa regulation reveals a novel genetic switch in which extracellular ligands control the status of critical nuclear cofactor complexes.

Nuclear receptor repression mediated by a complex containing SMRT, mSin3A, and histone deacetylase

The transcriptional corepressors SMRT and N-CoR function as silencing mediators for retinoid and thyroid hormone receptors. Here we show that SMRT and N-CoR directly interact with mSin3A, a corepressor for the Mad-Max heterodimer and a homolog of the yeast global-transcriptional repressor Sin3p. In addition, we demonstrate that the recently characterized histone deacetylase 1 (HDAC1) interacts with Sin3A and SMRT to form a multisubunit repressor complex. Consistent with this model, we find that HDAC inhibitors synergize with retinoic acid to stimulate hormone-responsive genes and differentiation of myeloid leukemia (HL-60) cells. This work establishes a convergence of repression pathways for bHLH-Zip proteins and nuclear receptors and suggests this type of regulation may be more widely conserved than previously suspected.

Identification of Prp40, a novel essential yeast splicing factor associated with the U1 small nuclear ribonucleoprotein particle

We have used suppressor genetics to identify factors that interact with Saccharomyces cerevisiae U1 small nuclear RNA (snRNA). In this way, we isolated PRP40-1, a suppressor that restores growth at 18 degrees C to a strain bearing a cold-sensitive mutation in U1 RNA. A gene disruption experiment shows that PRP40 is an essential gene. To study the role of PRP40 in splicing, we created a pool of temperature-sensitive prp40 strains. Primer extension analysis of intron-containing transcripts in prp40 temperature-sensitive strains reveals a splicing defect, indicating that Prp40 plays a direct role in pre-mRNA splicing. In addition, U1 RNA coimmunoprecipitates with Pro40, indicating that Prp40 is bound to the U1 small nuclear ribonucleoprotein particle in vivo. Therefore, we conclude that PRP40 encodes a novel, essential splicing component that associates with the yeast U1 small nuclear ribonucleoprotein particle.

The amino-terminal domain of yeast U1-70K is necessary and sufficient for function

The Saccharomyces cerevisiae SNP1 gene encodes a protein that shares 30% amino acid identity with the mammalian U1 small nuclear ribonucleoprotein particle protein 70K (U1-70K). We have demonstrated that yeast strains in which the SNP1 gene was disrupted are viable but exhibit greatly increased doubling times and severe temperature sensitivity. Furthermore, snp1-null strains are defective in pre-mRNA splicing. We have tested deletion alleles of SNP1 for their ability to complement these phenotypes. We found that the highly conserved RNA recognition motif consensus domain of Snp1 is not required for complementation of the snp1-null growth or splicing defects nor for the in vivo association with the U1 small nuclear ribonucleoprotein particle. However, the amino-terminal domain of Snp1, less strongly conserved, is necessary and sufficient for complementation.

Functional association of essential splicing factor(s) with PRP19 in a protein complex

We have previously shown that the yeast PRP19 protein is a spliceosomal component, but is not tightly associated with small nuclear RNAs. It appears to associate with the spliceosome concomitant with or just after dissociation of the U4 small nuclear RNA during spliceosome assembly. We have found that PRP19 is associated with a protein complex in the splicing extract and that at least one of the associated components is essential for splicing. Taking advantage of the epitope tagging technique, we have isolated the PRP19-associated complex by affinity chromatography. The isolated complex is functional for complementation for the heat-inactivated prp19 mutant extract, and consists of at least seven polypeptides in addition to PRP19. At least three of these can interact directly with the PRP19 protein. We also show that the PRP19 protein itself is in an oligomeric form, which might be a prerequisite for its interaction with these proteins.

The yeast homolog of the U1 snRNP protein 70K is encoded by the SNP1 gene

The product of the yeast SNP1 gene has high homology to two domains of the metazoan U1 snRNP protein 70K, which binds to stem/loop I of the U1 RNA. However, the absence of other domains conserved in metazoan 70K and the minimal effect of yeast U1 RNA stem/loop I deletion make the assignment of SNP1 as yeast 70K less clear. To address this question, we have expressed the SNP1 gene as a fusion protein in E. coli and developed a gel shift assay for U1 RNA binding. We show here that the product of the yeast SNP1 gene binds directly and specifically to the first 47 nucleotides of yeast U1 RNA, which include the stem/loop 1 structure. We therefore conclude that the SNP1 gene product is the yeast 70K homolog. This is the first yeast protein to be identified as a homolog of a metazoan snRNP protein.