Search

In Escherichia coli, acyl carrier protein (ACP) is posttranslationally converted into its active holo-ACP form via covalent linkage of 4'-phosphopantetheine (4'-PP) to residue serine-36. We found that the long flexible 4'-PP arm could react chemoselectively with the iodoacetyl group introduced on solid supports with high efficiency under mild conditions. Based on this finding, we developed site-selective immobilisation of proteins via the active holo-ACP fusion tag, independently of the physicochemical properties of the protein of interest. Furthermore, the molecular ratios of co-immobilised proteins can be manipulated because the tethering process is predominantly directed by the molar concentrations of diverse holo-ACP fusions during co-immobilisation. Conveniently tuning the molecular ratios of co-immobilised proteins allows their cooperation, leading to a highly productive multi-protein co-immobilisation system. Kinetic studies of enzymes demonstrated that α-amylase (Amy) and methyl parathion hydrolase (MPH) immobilised via active tag holo-ACP had higher catalytic efficiency (k

The Drosophila genome encodes three BEN-solo proteins including Insensitive (Insv), Elba1 and Elba2 that possess activities in transcriptional repression and chromatin insulation. A fourth protein—Elba3—bridges Elba1 and Elba2 to form an ELBA complex. Here, we report comprehensive investigation of these proteins in Drosophila embryos. We assess common and distinct binding sites for Insv and ELBA and their genetic interdependencies. While Elba1 and Elba2 binding generally requires the ELBA complex, Elba3 can associate with chromatin independently of Elba1 and Elba2. We further demonstrate that ELBA collaborates with other insulators to regulate developmental patterning. Finally, we find that adjacent gene pairs separated by an ELBA bound sequence become less differentially expressed in ELBA mutants. Transgenic reporters confirm the insulating activity of ELBA- and Insv-bound sites. These findings define ELBA and Insv as general insulator proteins in Drosophila and demonstrate the functional importance of insulators to partition transcription units., The BEN-solo proteins—including Insensitive (Insv), Elba1 and Elba2—function in both transcriptional repression and chromatin insulation. Here, the authors investigate the role of these proteins in Drosophila embryos, finding that ELBA and Insv function as general insulators and partition active chromatin to ensure proper gene activation in Drosophila.

In view of the non-interpretability of disease diagnosis models based on deep learning, a knowledge reasoning model based on medical knowledge graph for intelligent diagnosis is proposed. Given the patient symptom set, the co-occurrence of the patient and the disease is calculated, then the patient suffering from one disease is calculated. Based on the dynamic threshold value, the final disease diagnosis result of the patient is outputted. According to the symptoms of patients and the symptoms in the knowledge graph, the causal reasoning of the disease diagnosis is interpretable. Experiments on 145,712 pediatric electronic medical records in Chinese show that the proposed model can predict diseases with interpretability, and the accuracy reaches-82.12%.

Intent detection and slot filling are important modules in task-oriented dialog systems. In order to make full use of the relationship between different modules and resource sharing, solving the problem of a lack of semantics, this paper proposes a multitasking learning intent-detection system, based on the knowledge-base and slot-filling joint model. The approach has been used to share information and rich external utility between intent and slot modules in a three-part process. First, this model obtains shared parameters and features between the two modules based on long short-term memory and convolutional neural networks. Second, a knowledge base is introduced into the model to improve its performance. Finally, a weighted-loss function is built to optimize the joint model. Experimental results demonstrate that our model achieves better performance compared with state-of-the-art algorithms on a benchmark Airline Travel Information System (ATIS) dataset and the Snips dataset. Our joint model achieves state-of-the-art results on the benchmark ATIS dataset with a 1.33% intent-detection accuracy improvement, a 0.94% slot filling F value improvement, and with 0.19% and 0.31% improvements respectively on the Snips dataset.

Ureolytic microorganisms have attracted much interest for the bioprecipitation, removal and biorecovery of toxic metals from aqueous solution but little is known about the distribution of cationic/anionic components or quantification of metal species in a fungal system, or the contribution of extracellular protein structure and adsorption during precipitation. In this research, carbonate-rich fungal growth supernatants were applied for bioprecipitation of Ni-containing minerals. The solubility and stability of relevant nickel species was calculated and simulated using Geochemists' Workbench (GWB) and the structure and adsorption of extracellular protein were determined using attenuated total reflectance Fourier transform infrared spectroscopy (ATR-FTIR). The results showed that hydrated nickel carbonate and trace nickel phosphate minerals, with a proportion being of nanoscale dimensions, were precipitated after mixing fungal growth supernatant and NiCl2. The solubility diagrams calculated by GWB showed that under the same condition nickel phosphates were the only precipitate in the simulated system. X-ray photoelectron spectroscopy (XPS) results showed the existence of protein N in the minerals and a protein–nanoparticle bond in C 1s spectra, which showed that extracellular protein might be an important factor in the difference between experimental results and the geochemical simulation. ATR-FTIR showed that during the biomineralization process, the conformation of extracellular proteins changed and β structures increased at the expense of highly ordered α-helixes and other disordered structures, which might provide more nucleation sites to promote the crystallization of nickel carbonate especially under saturated conditions of CO32−. These results demonstrate that the combination of geochemical modelling and experimental analysis provide more complete scientific data to the explain differences between experimental results and theoretical outcomes. This can therefore enable reasonable predictions as to the effect of metal-containing nanoparticles on protein structure, and vice versa, during the process of metal bioprecipitation mediated by a carbonate-rich fungal growth supernatant system.

We present two new neighbor query algorithms, including range query (RNN) and nearest neighbor (NN) query, based on revised k-d tree by using two techniques. The first technique is proposed for decreasing unnecessary distance computations by checking whether the cell of a node is inside or outside the specified neighborhood of query point, and the other is used to reduce redundant visiting nodes by saving the indices of descendant points. We also implement the proposed algorithms in Matlab and C. The Matlab version is to improve original RNN and NN which are based on k-d tree, C version is to improve k-Nearest neighbor query (kNN) which is based on buffer k-d tree. Theoretical and experimental analysis have shown that the proposed algorithms significantly improve the original RNN, NN and kNN in low dimension, respectively. The tradeoff is that the additional space cost of the revised k-d tree is approximately O(αnlog (n)).

Representation Learning (RL) aims to convert data into low-dimensional and dense real-valued vectors, so as to realize reasoning in vector space. RL is one of the important research contents in the analysis of health data. This paper systematically reviews the latest research on Electronic Health Records (EHR) RL. We searched the Web of Science, Google Scholar, and Association for Computing Machinery Digital Library for papers involving EHR RL. On the basis of literature review, we propose a new taxonomy to categorize the state-of-the-art EHR RL methods into three categories: statistics learning-based RL methods, knowledge RL methods and graph RL methods. We analyze and summarize their characteristics according to the input data form and underlying learning mechanisms. In addition, we provide evaluation strategies to verify the quality of EHR representations from both intrinsic and extrinsic perspectives. Finally, we put forward three promising research directions to promote future research. Overall, this survey aims to provide a profound overview of state-of-the-art developments in the field of EHR RL and to help researchers find the most appropriate methods.

The goal of the question-answering (QA) system is to understand the questions from users and return their accurate answers. In the medical field, the question-answering system amis to understand patients' questions and return the correct answers. The existed knowledge base question-answering (KB-QA) systems mainly rely on hand-crafted features and ignore structure information of knowledge base which accordingly lead to the answers with low accuracy. In this paper, a novel KB-QA model is put forward based on knowledge representation and recurrent convolutional neural network. This model has three parts, candidate answers generation, entity relationships extraction and knowledge representation learning based on knowledge base. In addition, an algorithm is also developed to compute the scores of linking candidate answers and knowledge base. Experimental results show that the presented model achieves better performance compared with the baseline systems.

Distribution of datasets with respect to four types of classification applications for pediatric Chinese EMRs. (PDF 142 kb)

Background Benefiting from big data, powerful computation and new algorithmic techniques, we have been witnessing the renaissance of deep learning, particularly the combination of natural language processing (NLP) and deep neural networks. The advent of electronic medical records (EMRs) has not only changed the format of medical records but also helped users to obtain information faster. However, there are many challenges regarding researching directly using Chinese EMRs, such as low quality, huge quantity, imbalance, semi-structure and non-structure, particularly the high density of the Chinese language compared with English. Therefore, effective word segmentation, word representation and model architecture are the core technologies in the literature on Chinese EMRs. Results In this paper, we propose a deep learning framework to study intelligent diagnosis using Chinese EMR data, which incorporates a convolutional neural network (CNN) into an EMR classification application. The novelty of this paper is reflected in the following: (1) We construct a pediatric medical dictionary based on Chinese EMRs. (2) Word2vec adopted in word embedding is used to achieve the semantic description of the content of Chinese EMRs. (3) A fine-tuning CNN model is constructed to feed the pediatric diagnosis with Chinese EMR data. Our results on real-world pediatric Chinese EMRs demonstrate that the average accuracy and F1-score of the CNN models are up to 81%, which indicates the effectiveness of the CNN model for the classification of EMRs. Particularly, a fine-tuning one-layer CNN performs best among all CNNs, recurrent neural network (RNN) (long short-term memory, gated recurrent unit) and CNN-RNN models, and the average accuracy and F1-score are both up to 83%. Conclusion The CNN framework that includes word segmentation, word embedding and model training can serve as an intelligent auxiliary diagnosis tool for pediatricians. Particularly, a fine-tuning one-layer CNN performs well, which indicates that word order does not appear to have a useful effect on our Chinese EMRs. Electronic supplementary material The online version of this article (10.1186/s12859-019-2617-8) contains supplementary material, which is available to authorized users.

National Natural Science Foundation of China [61202144, 61203282]; Natural Science Foundation of Fujian Province in China [2012J01274, 2012J05125]; Key Laboratory of System Control and Information Processing; Ministry of Education of Shanghai Jiao Tong University [SCIP2012007]; Research Grant Council of Huaqiao University [09BS515]

Background The stability of Variable Importance Measures (VIMs) based on random forest has recently received increased attention. Despite the extensive attention on traditional stability of data perturbations or parameter variations, few studies include influences coming from the intrinsic randomness in generating VIMs, i.e. bagging, randomization and permutation. To address these influences, in this paper we introduce a new concept of intrinsic stability of VIMs, which is defined as the self-consistence among feature rankings in repeated runs of VIMs without data perturbations and parameter variations. Two widely used VIMs, i.e., Mean Decrease Accuracy (MDA) and Mean Decrease Gini (MDG) are comprehensively investigated. The motivation of this study is two-fold. First, we empirically verify the prevalence of intrinsic stability of VIMs over many real-world datasets to highlight that the instability of VIMs does not originate exclusively from data perturbations or parameter variations, but also stems from the intrinsic randomness of VIMs. Second, through Spearman and Pearson tests we comprehensively investigate how different factors influence the intrinsic stability. Results The experiments are carried out on 19 benchmark datasets with diverse characteristics, including 10 high-dimensional and small-sample gene expression datasets. Experimental results demonstrate the prevalence of intrinsic stability of VIMs. Spearman and Pearson tests on the correlations between intrinsic stability and different factors show that #feature (number of features) and #sample (size of sample) have a coupling effect on the intrinsic stability. The synthetic indictor, #feature/#sample, shows both negative monotonic correlation and negative linear correlation with the intrinsic stability, while OOB accuracy has monotonic correlations with intrinsic stability. This indicates that high-dimensional, small-sample and high complexity datasets may suffer more from intrinsic instability of VIMs. Furthermore, with respect to parameter settings of random forest, a large number of trees is preferred. No significant correlations can be seen between intrinsic stability and other factors. Finally, the magnitude of intrinsic stability is always smaller than that of traditional stability. Conclusion First, the prevalence of intrinsic stability of VIMs demonstrates that the instability of VIMs not only comes from data perturbations or parameter variations, but also stems from the intrinsic randomness of VIMs. This finding gives a better understanding of VIM stability, and may help reduce the instability of VIMs. Second, by investigating the potential factors of intrinsic stability, users would be more aware of the risks and hence more careful when using VIMs, especially on high-dimensional, small-sample and high complexity datasets.

The property of calibration of Multi-Label Learning (MLL) has not been well studied. Because of the excellent calibration property of Conformal Predictors (CP), it is valuable to achieve calibrated MLL prediction via CP. Three practical implementations of Multi-Label Conformal Predictors (MLCP) can be established. Among them are Instance Reproduction MLCP (IR-MLCP), Binary Relevance MLCP (BR-MLCP) and Power Set MLCP (PS-MLCP). The experimental results on benchmark datasets show that all three MLCP methods possess calibration property. Comparatively speaking, BR-MLCP performs better in terms of prediction efficiency and computational cost than the other two.

The recently developed Conformal Predictor (CP) can provide calibrated confidence for prediction which is out of the traditional predictors’ capacity. However, CP works for balanced data and fails in the case of imbalanced data. To handle this problem, Local Clustering Conformal Predictor (LCCP) which plugs a two-level partition into the framework of CP is proposed. In the first-level partition, the whole imbalanced training dataset is partitioned into some class-taxonomy data subsets. Secondly, the majority class examples proceed to be partitioned into some cluster-taxonomy data subsets by clustering method. To predict a new instance, LCCP selects the nearest cluster, incorporated with the minority class examples, to build a re-balanced training data. The designed LCCP model aims to not only provide valid confidence for prediction, but significantly improve the prediction efficiency as well. The experimental results show that LCCP model presents superiority than CP model for imbalanced data classification.

In order to improve the computational efficiency of conformal predictor, distance metric learning methods were used in the algorithm. The process of learning was divided into two stages: offline learning and online learning. Firstly, part of the training data was used in distance metric learning to get a space transformation matrix in the offline learning stage; Secondly, standard CP-KNN was conducted on the remaining training data with a nonconformity measure function defined by K nearest neighbors classifier in the transformed space. Experimental results on three UCI datasets demonstrate the efficiency of the new algorithm.

Many machine learning approaches have been proposed to establish the chronic gastritis diagnostic models. But till now, most of the machine-learning classifiers do not give any insight as to which features play key roles with respect to the derived classifier as well as the individual class. Recently, the variables importance measure yielded by random forest (RF) has been proposed in many applications. However, in multi-label classifications RF attempts to yield a common feature ranking for all classes, which fail in identifying the distinct predictive structures for individual class. This paper developed an improved random forest variables importance measure to evaluate the importance of features according to each individual class in multi-classification problem, and then applied a wrapper method for feature selection to construct the key features sets referring to each subtype of the chronic gastritis. Experiment results show that, compared with the previous studies, the selected features are more close to expert knowledge and contribute to better understanding of the underlying process that characterize the chronic gastritis.

Most of state-of-the-art machine learning algorithms cannot provide a reliable measure of their classifications and predictions. This paper addresses the importance of reliability and confidence for classification, and presents a novel method based on a combination of the unexcelled ensemble method, random forest (RF), and transductive confidence machine (TCM) which we call TCM-RF. The new algorithm hedges the predictions of RF and gives a well-calibrated region prediction by using the proximity matrix generated with RF as a nonconformity measure of examples. The new method takes advantage of RF and possesses a more precise and robust nonconformity measure. It can deal with redundant and noisy data with mixed types of variables, and is less sensitive to parameter settings. Experiments on benchmark datasets show it is more effective and robust than other TCMs. Further study on a real-world lymphoma microarray dataset shows its superiority over SVM with the ability of controlling the risk of error.

The output of traditional classifier is point prediction without giving any confidence of it. To the contrary, transductive confidence machine (TCM), which is a novel framework that provides a prediction result coupled with its accurate confidence. This method also can hedge the prediction in which the predicting accuracy will be controlled by predefined confidence level. In the framework of TCM, the efficiency of prediction depends on the strangeness function of samples. This paper incorporates random forests (RF) into the framework of TCM and proposes new TCM algorithm named TCM-RF, in which the strangeness obtained by RF will be used to implement the confidence prediction. Compared with traditional TCM algorithms, our method benefits from the more precise and robust strangeness measure and takes advantage of random forest. Experiments indicate its effectiveness and robustness. In addition, our study demonstrated that using ensemble strategies to define sample strangeness may be a more principled way than using a single classifier. On the other hand, it also shows that the paradigm of hedging prediction can be applied to an ensemble classifier.

Background Most machine-learning classifiers output label predictions for new instances without indicating how reliable the predictions are. The applicability of these classifiers is limited in critical domains where incorrect predictions have serious consequences, like medical diagnosis. Further, the default assumption of equal misclassification costs is most likely violated in medical diagnosis. Results In this paper, we present a modified random forest classifier which is incorporated into the conformal predictor scheme. A conformal predictor is a transductive learning scheme, using Kolmogorov complexity to test the randomness of a particular sample with respect to the training sets. Our method show well-calibrated property that the performance can be set prior to classification and the accurate rate is exactly equal to the predefined confidence level. Further, to address the cost sensitive problem, we extend our method to a label-conditional predictor which takes into account different costs for misclassifications in different class and allows different confidence level to be specified for each class. Intensive experiments on benchmark datasets and real world applications show the resultant classifier is well-calibrated and able to control the specific risk of different class. Conclusion The method of using RF outlier measure to design a nonconformity measure benefits the resultant predictor. Further, a label-conditional classifier is developed and turn to be an alternative approach to the cost sensitive learning problem that relies on label-wise predefined confidence level. The target of minimizing the risk of misclassification is achieved by specifying the different confidence level for different class.