ACM-BCB ... ... : the ... ACM Conference on Bioinformatics, Computational Biology and Biomedicine. ACM Conference on Bioinformatics, Computational Biology and Biomedicine最新文献

Working with electronic health records (EHRs) is known to be challenging due to several reasons. These reasons include not having: 1) similar lengths (per visit), 2) the same number of observations (per patient), and 3) complete entries in the available records. These issues hinder the performance of the predictive models created using EHRs. In this paper, we approach these issues by presenting a model for the combined task of imputing and predicting values for the irregularly observed and varying length EHR data with missing entries. Our proposed model (dubbed as Bi-GAN) uses a bidirectional recurrent network in a generative adversarial setting. In this architecture, the generator is a bidirectional recurrent network that receives the EHR data and imputes the existing missing values. The discriminator attempts to discriminate between the actual and the imputed values generated by the generator. Using the input data in its entirety, Bi-GAN learns how to impute missing elements in-between (imputation) or outside of the input time steps (prediction). Our method has three advantages to the state-of-the-art methods in the field: (a) one single model performs both the imputation and prediction tasks; (b) the model can perform predictions using time-series of varying length with missing data; (c) it does not require to know the observation and prediction time window during training and can be used for the predictions with different observation and prediction window lengths, for short- and long-term predictions. We evaluate our model on two large EHR datasets to impute and predict body mass index (BMI) values and show its superior performance in both settings.

Named entity recognition (NER) and normalization (EN) form an indispensable first step to many biomedical natural language processing applications. In biomedical information science, recognizing entities (e.g., genes, diseases, or drugs) and normalizing them to concepts in standard terminologies or thesauri (e.g., Entrez, ICD-10, or RxNorm) is crucial for identifying more informative relations among them that drive disease etiology, progression, and treatment. In this effort we pursue two high level strategies to improve biomedical ER and EN. The first is to decouple standard entity encoding tags (e.g., "B-Drug" for the beginning of a drug) into type tags (e.g., "Drug") and positional tags (e.g., "B"). A second strategy is to use additional counterfactual training examples to handle the issue of models learning spurious correlations between surrounding context and normalized concepts in training data. We conduct elaborate experiments using the MedMentions dataset, the largest dataset of its kind for ER and EN in biomedicine. We find that our first strategy performs better in entity normalization when compared with the standard coding scheme. The second data augmentation strategy uniformly improves performance in span detection, typing, and normalization. The gains from counterfactual examples are more prominent when evaluating in zero-shot settings, for concepts that have never been encountered during training.

The rapid adoption of electronic health records (EHRs) systems has made clinical data available in electronic format for research and for many downstream applications. Electronic screening of potentially eligible patients using these clinical databases for clinical trials is a critical need to improve trial recruitment efficiency. Nevertheless, manually translating free-text eligibility criteria into database queries is labor intensive and inefficient. To facilitate automated screening, free-text eligibility criteria must be structured and coded into a computable format using controlled vocabularies. Named entity recognition (NER) is thus an important first step. In this study, we evaluate 4 state-of-the-art transformer-based NER models on two publicly available annotated corpora of eligibility criteria released by Columbia University (i.e., the Chia data) and Facebook Research (i.e.the FRD data). Four transformer-based models (i.e., BERT, ALBERT, RoBERTa, and ELECTRA) pretrained with general English domain corpora vs. those pretrained with PubMed citations, clinical notes from the MIMIC-III dataset and eligibility criteria extracted from all the clinical trials on ClinicalTrials.gov were compared. Experimental results show that RoBERTa pretrained with MIMIC-III clinical notes and eligibility criteria yielded the highest strict and relaxed F-scores in both the Chia data (i.e., 0.658/0.798) and the FRD data (i.e., 0.785/0.916). With promising NER results, further investigations on building a reliable natural language processing (NLP)-assisted pipeline for automated electronic screening are needed.

With the rapid accumulation of electronic health record (EHR) data, deep learning (DL) models have exhibited promising performance on patient risk prediction. Recent advances have also demonstrated the effectiveness of knowledge graphs (KG) in providing valuable prior knowledge for further improving DL model performance. However, it is still unclear how KG can be utilized to encode high-order relations among clinical concepts and how DL models can make full use of the encoded concept relations to solve real-world healthcare problems and to interpret the outcomes. We propose a novel knowledge graph guided double attention LSTM model named KGDAL for rolling mortality prediction for critically ill patients with acute kidney injury requiring dialysis (AKI-D). KGDAL constructs a KG-based two-dimension attention in both time and feature spaces. In the experiment with two large healthcare datasets, we compared KGDAL with a variety of rolling mortality prediction models and conducted an ablation study to test the effectiveness, efficacy, and contribution of different attention mechanisms. The results showed that KGDAL clearly outperformed all the compared models. Also, KGDAL-derived patient risk trajectories may assist healthcare providers to make timely decisions and actions. The source code, sample data, and manual of KGDAL are available at https://github.com/lucasliu0928/KGDAL.

Integrating single-cell measurements that capture different properties of the genome is vital to extending our understanding of genome biology. This task is challenging due to the lack of a shared axis across datasets obtained from different types of single-cell experiments. For most such datasets, we lack corresponding information among the cells (samples) and the measurements (features). In this scenario, unsupervised algorithms that are capable of aligning single-cell experiments are critical to learning an in silico co-assay that can help draw correspondences among the cells. Maximum mean discrepancy-based manifold alignment (MMD-MA) is such an unsupervised algorithm. Without requiring correspondence information, it can align single-cell datasets from different modalities in a common shared latent space, showing promising results on simulations and a small-scale single-cell experiment with 61 cells. However, it is essential to explore the applicability of this method to larger single-cell experiments with thousands of cells so that it can be of practical interest to the community. In this paper, we apply MMD-MA to two recent datasets that measure transcriptome and chromatin accessibility in ~2000 single cells. To scale the runtime of MMD-MA to a more substantial number of cells, we extend the original implementation to run on GPUs. We also introduce a method to automatically select one of the user-defined parameters, thus reducing the hyperparameter search space. We demonstrate that the proposed extensions allow MMD-MA to accurately align state-of-the-art single-cell experiments.

Deep learning has shown a great promise in classifying brain disorders due to its powerful ability in learning optimal features by nonlinear transformation. However, given the high-dimension property of neuroimaging data, how to jointly exploit complementary information from multimodal neuroimaging data in deep learning is difficult. In this paper, we propose a novel multilevel convolutional neural network (CNN) fusion method that can effectively combine different types of neuroimage-derived features. Importantly, we incorporate a sequential feature selection into the CNN model to increase the feature interpretability. To evaluate our method, we classified two symptom-related brain disorders using large-sample multi-site data from 335 schizophrenia (SZ) patients and 380 autism spectrum disorder (ASD) patients within a cross-validation procedure. Brain functional networks, functional network connectivity, and brain structural morphology were employed to provide possible features. As expected, our fusion method outperformed the CNN model using only single type of features, as our method yielded higher classification accuracy (with mean accuracy >85%) and was more reliable across multiple runs in differentiating the two groups. We found that the default mode, cognitive control, and subcortical regions contributed more in their distinction. Taken together, our method provides an effective means to fuse multimodal features for the diagnosis of different psychiatric and neurological disorders.

Cryo-electron microscopy is a major structure determination technique for large molecular machines and membrane-associated complexes. Although atomic structures have been determined directly from cryo-EM density maps with high resolutions, current structure determination methods for medium resolution (5 to 10 Å) cryo-EM maps are limited by the availability of structure templates. Secondary structure traces are lines detected from a cryo-EM density map for α-helices and β-strands of a protein. When combined with secondary structure sequence segments predicted from a protein sequence, it is possible to generate a set of likely topologies of α-traces and β-sheet traces. A topology describes the overall folding relationship among secondary structures; it is a critical piece of information for deriving the corresponding atomic structure. We propose a method for protein structure prediction that combines three sources of information: the secondary structure traces detected from the cryo-EM density map, predicted secondary structure sequence segments, and amino acid contact pairs predicted using MULTICOM. A case study shows that using amino acid contact prediction from MULTICOM improves the ranking of the true topology. Our observations convey that using a small set of highly voted secondary structure contact pairs enhances the ranking in all experiments conducted for this case.

Although Cryo-electron microscopy (cryo-EM) has been successfully used to derive atomic structures for many proteins, it is still challenging to derive atomic structure when the resolution of cryo-EM density maps is in the medium range, e.g., 5-10 Å. Studies have attempted to utilize machine learning methods, especially deep neural networks to build predictive models for the detection of protein secondary structures from cryo-EM images, which ultimately helps to derive the atomic structure of proteins. However, the large variation in data quality makes it challenging to train a deep neural network with high prediction accuracy. Curriculum learning has been shown as an effective learning paradigm in machine learning. In this paper, we present a study using curriculum learning as a more effective way to utilize cryo-EM density maps with varying quality. We investigated three distinct training curricula that differ in whether/how images used for training in past are reused while the network was continually trained using new images. A total of 1,382 3-dimensional cryo-EM images were extracted from density maps of Electron Microscopy Data Bank in our study. Our results indicate learning with curriculum significantly improves the performance of the final trained network when the forgetting problem is properly addressed.

Single cell RNA sequencing is a powerful technique that measures the gene expression of individual cells in a high throughput fashion. However, due to sequencing inefficiency, the data is unreliable due to dropout events, or technical artifacts where genes erroneously appear to have zero expression. Many data imputation methods have been proposed to alleviate this issue. Yet, effective imputation can be difficult and biased because the data is sparse and high-dimensional, resulting in major distortions in downstream analyses. In this paper, we propose a completely novel approach that imputes the gene-by-gene correlations rather than the data itself. We call this method SCENA: Single cell RNA-seq Correlation completion by ENsemble learning and Auxiliary information. The SCENA gene-by-gene correlation matrix estimate is obtained by model stacking of multiple imputed correlation matrices based on known auxiliary information about gene connections. In an extensive simulation study based on real scRNA-seq data, we demonstrate that SCENA not only accurately imputes gene correlations but also outperforms existing imputation approaches in downstream analyses such as dimension reduction, cell clustering, graphical model estimation.