Biology subjects

Xie, X.

Publications and source records attributed to Xie, X..

260 records · Page 15Linked to original sources

bioRxiv2017

DeepLung: 3D Deep Convolutional Nets for Automated Pulmonary Nodule Detection and Classification

In this work, we present a fully automated lung CT cancer diagnosis system, DeepLung. DeepLung contains two parts, nodule detection and classification. Considering the 3D nature of lung CT data, two 3D networks are designed for the nodule detection and classification respectively. Specifically, a 3D Faster R-CNN is designed for nodule detection with a U-net-like encoder-decoder structure to effectively learn nodule features. For nodule classification, gradient boosting machine (GBM) with 3D dual path network (DPN) features is proposed. The nodule classification subnetwork is validated on a public dataset from LIDC-IDRI, on which it achieves better performance than state-of-the-art approaches, and surpasses the average performance of four experienced doctors. For the DeepLung system, candidate nodules are detected first by the nodule detection subnetwork, and nodule diagnosis is conducted by the classification subnetwork. Extensive experimental results demonstrate the DeepLung is comparable to the experienced doctors both for the nodule-level and patient-level diagnosis on the LIDC-IDRI dataset.

bioinformatics↗

bioRxiv2017

FactorNet: a deep learning framework for predicting cell type specific transcription factor binding from nucleotide-resolution sequential data

Due to the large numbers of transcription factors (TFs) and cell types, querying binding profiles of all TF/cell type pairs is not experimentally feasible, owing to constraints in time and resources. To address this issue, we developed a convolutional-recurrent neural network model, called FactorNet, to computationally impute the missing binding data. FactorNet trains on binding data from reference cell types to make accurate predictions on testing cell types by leveraging a variety of features, including genomic sequences, genome annotations, gene expression, and single-nucleotide resolution sequential signals, such as DNase I cleavage. To the best of our knowledge, this is the first deep learning method to study the rules governing TF binding at such a fine resolution. With FactorNet, a researcher can perform a single sequencing assay, such as DNase-seq, on a cell type and computationally impute dozens of TF binding profiles. This is an integral step for reconstructing the complex networks underlying gene regulation. While neural networks can be computationally expensive to train, we introduce several novel strategies to significantly reduce the overhead. By visualizing the neural network models, we can interpret how the model predicts binding which in turn reveals additional insights into regulatory grammar. We also investigate the variables that affect cross-cell type predictive performance to explain why the model performs better on some TF/cell types than others, and offer insights to improve upon this field. Our method ranked among the top four teams in the ENCODE-DREAM in vivo Transcription Factor Binding Site Prediction Challenge.

genomics↗

bioRxiv2017

The Effect Of Nipped-B-Like (Nipbl) Haploinsufficiency On Genome-Wide Cohesin Binding And Target Gene Expression: Modeling Cornelia de Lange Syndrome

Cornelia de Lange Syndrome (CdLS) is a multisystem developmental disorder frequently associated with heterozygous loss-of-function mutations of Nipped-B-like (NIPBL), the human homolog of Drosophila Nipped-B. NIPBL loads cohesin onto chromatin. Cohesin mediates sister chromatid cohesion important for mitosis, but is also increasingly recognized as a regulator of gene expression. In CdLS patient cells and animal models, the presence of multiple gene expression changes with little or no sister chromatid cohesion defect suggests that disruption of gene regulation underlies this disorder. However, the effect of NIPBL haploinsufficiency on cohesin binding, and how this relates to the clinical presentation of CdLS, has not been fully investigated. Nipbl haploinsufficiency causes CdLS-like phenotype in mice. We examined genome-wide cohesin binding and its relationship to gene expression using mouse embryonic fibroblasts (MEFs) from Nipbl +/- mice that recapitulate the CdLS phenotype. We found a global decrease in cohesin binding, including at CCCTC-binding factor (CTCF) binding sites and repeat regions. Cohesin-bound genes were found to be enriched for histone H3 lysine 4 trimethylation (H3K4me3) at their promoters; were disproportionately downregulated in Nipbl mutant MEFs; and displayed evidence of reduced promoter-enhancer interaction. The results suggest that gene activation is the primary cohesin function sensitive to Nipbl reduction. Over 50% of significantly dysregulated transcripts in mutant MEFs come from cohesin target genes, including genes involved in adipogenesis that have been implicated in contributing to the CdLS phenotype. Thus, decreased cohesin binding at the gene regions directly contributes to disease-specific expression changes. Taken together, our Nipbl haploinsufficiency model allows us to analyze the dosage effect of cohesin loading on CdLS development.

developmental biology↗

bioRxiv2017

miR-200c Suppresses Stemness And Increases Cellular Sensitivity To Trastuzumab In HER2+ Breast Cancer

Resistance to trastuzumab remains a major obstacle in HER2-overexpressing breast cancer treatment. miR-200c is important for many functions in cancer stem cells (CSCs), including tumor recurrence, metastasis and resistance. We hypothesized that miR-200c contributes to trastuzumab resistance and stemness maintenance in HER2-overexpressing breast cancer. In this study, we used HER2-positive SKBR3, HER2-negative MCF-7, and their CD44+CD24- phenotype mammospheres SKBR3-S and MCF-7-S to verify. Our results demonstrated that miR-200c was weakly expressed in breast cancer cell lines and cell line stem cells. Overexpression of miR-200c resulted in a significant reduction in the number of tumor spheres formed and the population of CD44+CD24- phenotype mammospheres in SKBR3-S. Combining miR-200c with trastuzumab can significantly reduce proliferation and increase apoptosis of SKBR3 and SKBR3-S. Overexpression of miR-200c also eliminated its downstream target genes. These genes were highly expressed and positively related in breast cancer patients. Overexpression of miR-200c also improved the malignant progression of SKBR3-S and SKBR3 in vivo. miR-200c plays an important role in the maintenance of the CSC-like phenotype and increases drug sensitivity to trastuzumab in HER2+ cells and stem cells.\n\nSummary statementmiRNAs are critical in stemness maintenance and drug resistance. These data link maintenance of the stemness-related phenotype and the sensitivity of HER2+ breast cancer to miR-200c in response to trastuzumab.

cancer biology↗

bioRxiv2017

Understanding sequence conservation with deep learning

MotivationComparing the human genome to the genomes of closely related mammalian species has been a powerful tool for discovering functional elements in the human genome. Millions of conserved elements have been discovered. However, understanding the functional roles of these elements still remain a challenge, especially in noncoding regions. In particular, it is still unclear why these elements are evolutionarily conserved and what kind of functional elements are encoded within these sequences.\n\nResultsWe present a deep learning framework, called DeepCons, to uncover potential functional elements within conserved sequences. DeepCons is a convolutional neural net (CNN) that receives a short segment of DNA sequence as input and outputs the probability of the sequence of being evolutionary conserved. DeepCons utilizes hundreds of convolution kernels to detect features within DNA sequences, and automatically learns these kernels after training the CNN model using 887,577 conserved elements and a similar number of nonconserved elements in the human genome. On a balanced test dataset, DeepCons can achieve an accuracy of 75% in determining whether a sequence element is conserved or not, and the area under the ROC curve of 0.83, based on information from the human genome alone. We further investigate the properties of the learned kernels. Some kernels are directly related to well-known regulatory motifs corresponding to transcription factors. Many kernels show positional biases relative to transcriptional start sites or transcription end sites. But most of discovered kernels do not correspond to any known functional element, suggesting that they might represent unknown categories of functional elements. We also utilize DeepCons to annotate how changes at each individual nucleotide might impact the conservation properties of the surrounding sequences.\n\nAvailabilityThe source code of DeepCons and all the learned convolution kernels in motif format is publicly available online at https://github.com/uci-cbcl/DeepCons.\n\nContactxhx@ics.uci.edu

bioinformatics↗

bioRxiv2017

HLA class I binding prediction via convolutional neural networks

Many biological processes are governed by protein-ligand interactions. One such example is the recognition of self and non-self cells by the immune system. This immune response process is regulated by the major histocompatibility complex (MHC) protein which is encoded by the human leukocyte antigen (HLA) complex. Understanding the binding potential between MHC and peptides can lead to the design of more potent, peptide-based vaccines and immunotherapies for infectious autoimmune diseases.\n\nWe apply machine learning techniques from the natural language processing (NLP) domain to address the task of MHC-peptide binding prediction. More specifically, we introduce a new distributed representation of amino acids, name HLA-Vec, that can be used for a variety of downstream proteomic machine learning tasks. We then propose a deep convolutional neural network architecture, name HLA-CNN, for the task of HLA class I-peptide binding prediction. Experimental results show combining the new distributed representation with our HLA-CNN architecture acheives state-of-the-art results in the majority of the latest two Immune Epitope Database (IEDB) weekly automated benchmark datasets. We further apply our model to predict binding on the human genome and identify 15 genes with potential for self binding. Codes are available at https://github.com/uci-cbcl/HLA-bind.

bioinformatics↗

bioRxiv2016

Adversarial Deep Structural Networks for Mammographic Mass Segmentation

Mass segmentation is an important task in mammogram analysis, providing effective morphological features and regions of interest (ROI) for mass detection and classification. Inspired by the success of using deep convolutional features for natural image analysis and conditional random fields (CRF) for structural learning, we propose an end-to-end network for mammographic mass segmentation. The network employs a fully convolutional network (FCN) to model potential function, followed by a CRF to perform structural learning. Because the mass distribution varies greatly with pixel position, the FCN is combined with position priori for the task. Due to the small size of mammogram datasets, we use adversarial training to control over-fitting. Four models with different convolutional kernels are further fused to improve the segmentation results. Experimental results on two public datasets, INbreast and DDSM-BCRP, show that our end-to-end network combined with adversarial training achieves the-state-of-the-art results.

bioinformatics↗

bioRxiv2016

Deep Multi-instance Networks with Sparse Label Assignment for Whole Mammogram Classification

Mammogram classification is directly related to computer-aided diagnosis of breast cancer. Traditional methods requires great effort to annotate the training data by costly manual labeling and specialized computational models to detect these annotations during test. Inspired by the success of using deep convolutional features for natural image analysis and multi-instance learning for labeling a set of instances/patches, we propose end-to-end trained deep multiinstance networks for mass classification based on whole mammogram without the aforementioned costly need to annotate the training data. We explore three different schemes to construct deep multi-instance networks for whole mammogram classification. Experimental results on the INbreast dataset demonstrate the robustness of proposed deep networks compared to previous work using segmentation and detection annotations in the training.

bioinformatics↗