Pub Date : 2016-12-01DOI: 10.1109/BIBM.2016.7822511
Changsheng Zhang, Hongmin Cai, Jingying Huang, Bo Xu
The revolutionary invention of single-cell sequencing technology carves out a new way to delineate intra tumor heterogeneity and the evolution of single cells at the molecular level. Since single-cell sequencing requires a special genome amplification step to accumulate enough samples, a large number of bias were introduced, making the calling of copy number variants rather challenging. Accurately modeling this process and effectively detecting copy number variations (CNVs) are the major roadblock for single-cell sequencing data analysis. Recent advances manifested that the underlying copy numbers are corrupted by noise, which could be approximated by negative binomial distribution. In this paper, we formulated a general mathematical model for copy number reconstruction from read depth signal, and presented its two specific variants, namely Poisson-CNV and NB-CNV to catering for various reads distribution. Efficient numerical solution based on the classical alternating direction minimization method was designed to solve the proposed models. Extensive experiments on both synthetic datasets and empirical single-cell sequencing datasets were conducted to compare the performance of the two models. The results show that the proposed model of NB-CNV achieved superior performance in calling the CNV for single-cell sequencing data.
{"title":"Multi-norm constrained optimization methods for calling copy number variants in single cell sequencing data","authors":"Changsheng Zhang, Hongmin Cai, Jingying Huang, Bo Xu","doi":"10.1109/BIBM.2016.7822511","DOIUrl":"https://doi.org/10.1109/BIBM.2016.7822511","url":null,"abstract":"The revolutionary invention of single-cell sequencing technology carves out a new way to delineate intra tumor heterogeneity and the evolution of single cells at the molecular level. Since single-cell sequencing requires a special genome amplification step to accumulate enough samples, a large number of bias were introduced, making the calling of copy number variants rather challenging. Accurately modeling this process and effectively detecting copy number variations (CNVs) are the major roadblock for single-cell sequencing data analysis. Recent advances manifested that the underlying copy numbers are corrupted by noise, which could be approximated by negative binomial distribution. In this paper, we formulated a general mathematical model for copy number reconstruction from read depth signal, and presented its two specific variants, namely Poisson-CNV and NB-CNV to catering for various reads distribution. Efficient numerical solution based on the classical alternating direction minimization method was designed to solve the proposed models. Extensive experiments on both synthetic datasets and empirical single-cell sequencing datasets were conducted to compare the performance of the two models. The results show that the proposed model of NB-CNV achieved superior performance in calling the CNV for single-cell sequencing data.","PeriodicalId":345384,"journal":{"name":"2016 IEEE International Conference on Bioinformatics and Biomedicine (BIBM)","volume":"35 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2016-12-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"116640622","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2016-12-01DOI: 10.1109/BIBM.2016.7822673
M. Madkour, Hsing-yi Song, Jingcheng Du, Cui Tao
This paper describes a proposition for representing temporal indeterminacy in events from clinical narratives using fuzzy sets membership functions. This approach leverages both temporal and semantic information of events and has been proved by representational analysis evaluation method. We demonstrate that membership functions' graphs can be used for representing temporal approximation and granularity of events. We also show that this approach is helpful for the construction of fine timeline of clinical events, and can be used for calculating accurate metrics for ordering events.
{"title":"A representational analysis of a temporal indeterminancy display in clinical events","authors":"M. Madkour, Hsing-yi Song, Jingcheng Du, Cui Tao","doi":"10.1109/BIBM.2016.7822673","DOIUrl":"https://doi.org/10.1109/BIBM.2016.7822673","url":null,"abstract":"This paper describes a proposition for representing temporal indeterminacy in events from clinical narratives using fuzzy sets membership functions. This approach leverages both temporal and semantic information of events and has been proved by representational analysis evaluation method. We demonstrate that membership functions' graphs can be used for representing temporal approximation and granularity of events. We also show that this approach is helpful for the construction of fine timeline of clinical events, and can be used for calculating accurate metrics for ordering events.","PeriodicalId":345384,"journal":{"name":"2016 IEEE International Conference on Bioinformatics and Biomedicine (BIBM)","volume":"79 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2016-12-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"116688591","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2016-12-01DOI: 10.1109/BIBM.2016.7822585
Hao Jiang, W. Ching, Yushan Qiu, Xiaoqing Cheng
Positive semi-definiteness is a critical property in Support Vector Machine (SVM) methods to ensure efficient solutions through convex quadratic programming. In this paper, we introduce a projection matrix on indefinite kernels to formulate a positive semi-definite one. The proposed model can be regarded as a generalized version of the spectrum method (denoising method and flipping method) by varying parameter λ. In particular, our suggested optimal λ under the Bregman matrix divergence theory can be obtained using unconstrained optimization. Experimental results on 4 real world data sets ranging from glycan classification to cancer prediction show that the proposed model can achieve better or competitive performance when compared to the related indefinite kernel methods. This may suggest a new way in motif extractions or cancer predictions.
{"title":"Unconstrained optimization in projection method for indefinite SVMs","authors":"Hao Jiang, W. Ching, Yushan Qiu, Xiaoqing Cheng","doi":"10.1109/BIBM.2016.7822585","DOIUrl":"https://doi.org/10.1109/BIBM.2016.7822585","url":null,"abstract":"Positive semi-definiteness is a critical property in Support Vector Machine (SVM) methods to ensure efficient solutions through convex quadratic programming. In this paper, we introduce a projection matrix on indefinite kernels to formulate a positive semi-definite one. The proposed model can be regarded as a generalized version of the spectrum method (denoising method and flipping method) by varying parameter λ. In particular, our suggested optimal λ under the Bregman matrix divergence theory can be obtained using unconstrained optimization. Experimental results on 4 real world data sets ranging from glycan classification to cancer prediction show that the proposed model can achieve better or competitive performance when compared to the related indefinite kernel methods. This may suggest a new way in motif extractions or cancer predictions.","PeriodicalId":345384,"journal":{"name":"2016 IEEE International Conference on Bioinformatics and Biomedicine (BIBM)","volume":"4 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2016-12-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"116833785","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2016-12-01DOI: 10.1109/BIBM.2016.7822811
Hyojin Kang, Junehawk Lee, S. Yu
Differential Co-Expression Networks (DCENs) are graphical representations of genes showing differential co-expression pattern in response to experimental conditions or genetic changes. They have been successfully applied to identify condition-specific modules and provide a picture of the dynamic changes in gene regulatory networks. DCENs analysis investigates the differences among gene interconnections by calculating the expression correlation change of each gene pair between conditions. In this study, we collected many different datasets from NCBI GEO including 25 RNA-seq and 2,102 microarray samples derived from human brain and blood in Alzheimer's disease and performed differential co-expression analyses to identify functional modules responsible for the characterization of Alzheimer's disease. The DCENs were generated using Pearson correlation coefficient and meta-analysis was conducted using rank-based method. The preliminary results show that the structural characteristics of DCENs can provide new insights into the underlying gene regulatory dynamics in Alzheimer's disease. There is low size overlap between microarray- and RNA-seq-derived DCENs however, DCENs from RNA-seq would complement ones from microarray due to the higher coverage and dynamic range of RNA-seq.
{"title":"Differential Co-Expression Networks using RNA-seq and microarrays in Alzheimer's disease","authors":"Hyojin Kang, Junehawk Lee, S. Yu","doi":"10.1109/BIBM.2016.7822811","DOIUrl":"https://doi.org/10.1109/BIBM.2016.7822811","url":null,"abstract":"Differential Co-Expression Networks (DCENs) are graphical representations of genes showing differential co-expression pattern in response to experimental conditions or genetic changes. They have been successfully applied to identify condition-specific modules and provide a picture of the dynamic changes in gene regulatory networks. DCENs analysis investigates the differences among gene interconnections by calculating the expression correlation change of each gene pair between conditions. In this study, we collected many different datasets from NCBI GEO including 25 RNA-seq and 2,102 microarray samples derived from human brain and blood in Alzheimer's disease and performed differential co-expression analyses to identify functional modules responsible for the characterization of Alzheimer's disease. The DCENs were generated using Pearson correlation coefficient and meta-analysis was conducted using rank-based method. The preliminary results show that the structural characteristics of DCENs can provide new insights into the underlying gene regulatory dynamics in Alzheimer's disease. There is low size overlap between microarray- and RNA-seq-derived DCENs however, DCENs from RNA-seq would complement ones from microarray due to the higher coverage and dynamic range of RNA-seq.","PeriodicalId":345384,"journal":{"name":"2016 IEEE International Conference on Bioinformatics and Biomedicine (BIBM)","volume":"103 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2016-12-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"131161465","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2016-12-01DOI: 10.1109/BIBM.2016.7822665
Zhijun Yao, Bin Hu, Xuejiao Chen, Yuanwei Xie, Lei Fang
Recent studies suggested that cognitive impairments and memory difficulties in cancer survivors were associated with topology changes of brain network, particularly in terms of the functional and structural abnormalities. However, little is known about the modular reconfiguration of metabolic brain network among this population. In this study, we recruited 78 patients with pre-treatment cancer and 80 age- and gender-matched normal controls (NCs), and constructed the metabolic brain networks derived from resting-state 18F-fluorodeoxyglucose positron emission tomography (FDG-PET) to assess the alters of modularity pattern in cancer. The measurements of the participation index (PI) and mutual information (MI) were calculated for the cancer and NC groups. Compared with NC group, one module composed by the hippocampus, the amygdala and frontal and temporal regions was absented in cancer group. Moreover, cancer patients showed abnormal topology pattern in their metabolic networks (i.e., increased local efficiency and reduced global efficiency). Although node-wise PI shared positive correlated with normalized metabolism uptake in both groups, the more energy consumption were observed in metabolism network of cancer group that might be indicative of reduced capability of information processing. In addition, the between-group MIs were gradually increased over a range of thresholds. Our results suggested that modular pattern of the metabolic brain network seemed to re-shape its organization in cancer, which might uncover the neurobiological mechanisms underlying cancer-related cognitive dysfunction.
{"title":"Modular reconfiguration of metabolic brain networks in health and cancer: A resting-state PET study","authors":"Zhijun Yao, Bin Hu, Xuejiao Chen, Yuanwei Xie, Lei Fang","doi":"10.1109/BIBM.2016.7822665","DOIUrl":"https://doi.org/10.1109/BIBM.2016.7822665","url":null,"abstract":"Recent studies suggested that cognitive impairments and memory difficulties in cancer survivors were associated with topology changes of brain network, particularly in terms of the functional and structural abnormalities. However, little is known about the modular reconfiguration of metabolic brain network among this population. In this study, we recruited 78 patients with pre-treatment cancer and 80 age- and gender-matched normal controls (NCs), and constructed the metabolic brain networks derived from resting-state 18F-fluorodeoxyglucose positron emission tomography (FDG-PET) to assess the alters of modularity pattern in cancer. The measurements of the participation index (PI) and mutual information (MI) were calculated for the cancer and NC groups. Compared with NC group, one module composed by the hippocampus, the amygdala and frontal and temporal regions was absented in cancer group. Moreover, cancer patients showed abnormal topology pattern in their metabolic networks (i.e., increased local efficiency and reduced global efficiency). Although node-wise PI shared positive correlated with normalized metabolism uptake in both groups, the more energy consumption were observed in metabolism network of cancer group that might be indicative of reduced capability of information processing. In addition, the between-group MIs were gradually increased over a range of thresholds. Our results suggested that modular pattern of the metabolic brain network seemed to re-shape its organization in cancer, which might uncover the neurobiological mechanisms underlying cancer-related cognitive dysfunction.","PeriodicalId":345384,"journal":{"name":"2016 IEEE International Conference on Bioinformatics and Biomedicine (BIBM)","volume":"91 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2016-12-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"132394965","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2016-12-01DOI: 10.1109/BIBM.2016.7822744
Cunjin Luo, Kuanquan Wang, Henggui Zhang
The identified genetic short QT syndrome (SQTS) is associated with an increased risk of arrhythmia and sudden death. This study was to investigate the potential effects of propafenone on KCNH2-linked short QT syndrome (SQT1) using a multi-scale biophysically detailed model of the heart developed by ten Tusscher and Panfilov. The ion electrical conductivities were reduced by propafenone in order to simulate the pharmacological effects in healthy and SQT1 cells. Based on the experimental data of McPate et al., the pharmacological effect of propafenone was modelled by dose-dependent IKr blocking. Action potential (AP) profiles and 1D tissue level were analyzed to predict the effects of propafenone on SQT1. Both low- and high- dose of propafenone prolonged APD and QT interval in SQT1 cells. It suggests the superior efficacy of high dose of propafenone on SQT1. However, propafenone did not significantly alter the healthy APD or QT interval at low dose, whereas markedly shortened them at high dose. Our simulation data show that propafenone has a dose-dependently anti-arrhythmic effect on SQT1, and a pro-arrhythmic effect on healthy cells. These computer simulations help to better understand the underlying mechanisms responsible for the initiation or termination of arrhythmias in healthy or SQT1 patients using propafenone.
{"title":"Effects of propafenone on KCNH2-linked short QT syndrome: A modelling study","authors":"Cunjin Luo, Kuanquan Wang, Henggui Zhang","doi":"10.1109/BIBM.2016.7822744","DOIUrl":"https://doi.org/10.1109/BIBM.2016.7822744","url":null,"abstract":"The identified genetic short QT syndrome (SQTS) is associated with an increased risk of arrhythmia and sudden death. This study was to investigate the potential effects of propafenone on KCNH2-linked short QT syndrome (SQT1) using a multi-scale biophysically detailed model of the heart developed by ten Tusscher and Panfilov. The ion electrical conductivities were reduced by propafenone in order to simulate the pharmacological effects in healthy and SQT1 cells. Based on the experimental data of McPate et al., the pharmacological effect of propafenone was modelled by dose-dependent IKr blocking. Action potential (AP) profiles and 1D tissue level were analyzed to predict the effects of propafenone on SQT1. Both low- and high- dose of propafenone prolonged APD and QT interval in SQT1 cells. It suggests the superior efficacy of high dose of propafenone on SQT1. However, propafenone did not significantly alter the healthy APD or QT interval at low dose, whereas markedly shortened them at high dose. Our simulation data show that propafenone has a dose-dependently anti-arrhythmic effect on SQT1, and a pro-arrhythmic effect on healthy cells. These computer simulations help to better understand the underlying mechanisms responsible for the initiation or termination of arrhythmias in healthy or SQT1 patients using propafenone.","PeriodicalId":345384,"journal":{"name":"2016 IEEE International Conference on Bioinformatics and Biomedicine (BIBM)","volume":"4 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2016-12-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"132437700","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2016-12-01DOI: 10.1109/BIBM.2016.7822500
Keke Chen, Venkata Sai Abhishek Gogu, Di Wu, Jiang Ning
Lineage analysis has been an important method for understanding the mutation patterns and the diversity of genes, such as antibodies. A mutation lineage is typically represented as a tree structure, describing the possible mutation paths. Generating lineage trees from sequence data imposes two unique challenges: (1) Types of constraints might be defined on top of sequence data and tree structures, which have to be appropriately formulated and maintained by the algorithms. (2) Enumerating all possible trees that satisfy constraints is typically computationally intractable. In this paper, we present a COnstrained Lineage Tree generation framework (COLT) that builds lineage trees from sequences, based on local and global constraints specified by domain experts and heuristics derived from the mutation processes. Our formal analysis and experimental results show that this framework can efficiently generate valid lineage trees, while strictly satisfying the constraints specified by domain experts.
{"title":"COLT: COnstrained Lineage Tree Generation from sequence data","authors":"Keke Chen, Venkata Sai Abhishek Gogu, Di Wu, Jiang Ning","doi":"10.1109/BIBM.2016.7822500","DOIUrl":"https://doi.org/10.1109/BIBM.2016.7822500","url":null,"abstract":"Lineage analysis has been an important method for understanding the mutation patterns and the diversity of genes, such as antibodies. A mutation lineage is typically represented as a tree structure, describing the possible mutation paths. Generating lineage trees from sequence data imposes two unique challenges: (1) Types of constraints might be defined on top of sequence data and tree structures, which have to be appropriately formulated and maintained by the algorithms. (2) Enumerating all possible trees that satisfy constraints is typically computationally intractable. In this paper, we present a COnstrained Lineage Tree generation framework (COLT) that builds lineage trees from sequences, based on local and global constraints specified by domain experts and heuristics derived from the mutation processes. Our formal analysis and experimental results show that this framework can efficiently generate valid lineage trees, while strictly satisfying the constraints specified by domain experts.","PeriodicalId":345384,"journal":{"name":"2016 IEEE International Conference on Bioinformatics and Biomedicine (BIBM)","volume":"104 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2016-12-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"132477300","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2016-12-01DOI: 10.1109/BIBM.2016.7822663
P. Vizza, P. Guzzi, P. Veltri, G. Cascini, R. Curia, Loredana Sisca
The analysis of bioimages and their correlated clinical patient information allows to investigate specific diseases and define the corresponding medical protocols. To perform a correct diagnosis and apply a precise therapy, bioimages must be collected and studied together with others relevant data as well as laboratory results, medical annotations and patient history. Today, the management of these data is performed by single systems inside hospital departments that often do not provide dedicated data integration platforms among different departments as well as different health structures to exchange of relevant clinical information. Also, images cannot be annotated or enriched by physicians to trace temporal studies for patients or even among patients with similar diseases. In this contribution, we report the results of a research project called GIDAC (standing for Gestione Integrata DAti Clinici) that aims to define a general purpose framework for the bioimages management and annotations as well as clinical data view and integration in a simple-to-use information system. The proposed framework does not substitute any existing clinical information system but is able in gathering and integrating data by using a XML-based module. The novelty also consists in allowing annotations on DICOM images by means of simple user-interface to take trace of changes intra images as well as comparisons among patients. This system supports oncologists in the management of DICOM images from different devices (e.g., ecograph or PACS) to extract relevant information necessary to query (annotate) images and study similar clinical cases.
生物图像及其相关临床患者信息的分析允许调查特定疾病并确定相应的医疗方案。为了进行正确的诊断和应用精确的治疗，必须收集生物图像并与其他相关数据以及实验室结果、医学注释和患者病史一起研究。目前，这些数据的管理是由医院部门内部的单一系统完成的，这些系统往往没有在不同部门和不同医疗机构之间提供专用的数据集成平台来交换相关的临床信息。此外，医生无法对图像进行注释或丰富，以追踪患者甚至患有类似疾病的患者的时间研究。在这篇文章中，我们报告了一个名为GIDAC (Gestione Integrata DAti Clinici)的研究项目的结果，该项目旨在定义一个通用框架，用于生物图像管理和注释，以及临床数据视图和集成在一个简单易用的信息系统中。该框架不替代任何现有的临床信息系统，而是能够使用基于xml的模块收集和集成数据。其新颖之处还在于允许通过简单的用户界面对DICOM图像进行注释，以跟踪图像内的变化以及患者之间的比较。该系统支持肿瘤学家管理来自不同设备(如ecograph或PACS)的DICOM图像，以提取查询(注释)图像和研究类似临床病例所需的相关信息。
{"title":"GIDAC: A prototype for bioimages annotation and clinical data integration","authors":"P. Vizza, P. Guzzi, P. Veltri, G. Cascini, R. Curia, Loredana Sisca","doi":"10.1109/BIBM.2016.7822663","DOIUrl":"https://doi.org/10.1109/BIBM.2016.7822663","url":null,"abstract":"The analysis of bioimages and their correlated clinical patient information allows to investigate specific diseases and define the corresponding medical protocols. To perform a correct diagnosis and apply a precise therapy, bioimages must be collected and studied together with others relevant data as well as laboratory results, medical annotations and patient history. Today, the management of these data is performed by single systems inside hospital departments that often do not provide dedicated data integration platforms among different departments as well as different health structures to exchange of relevant clinical information. Also, images cannot be annotated or enriched by physicians to trace temporal studies for patients or even among patients with similar diseases. In this contribution, we report the results of a research project called GIDAC (standing for Gestione Integrata DAti Clinici) that aims to define a general purpose framework for the bioimages management and annotations as well as clinical data view and integration in a simple-to-use information system. The proposed framework does not substitute any existing clinical information system but is able in gathering and integrating data by using a XML-based module. The novelty also consists in allowing annotations on DICOM images by means of simple user-interface to take trace of changes intra images as well as comparisons among patients. This system supports oncologists in the management of DICOM images from different devices (e.g., ecograph or PACS) to extract relevant information necessary to query (annotate) images and study similar clinical cases.","PeriodicalId":345384,"journal":{"name":"2016 IEEE International Conference on Bioinformatics and Biomedicine (BIBM)","volume":"74 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2016-12-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"132232367","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2016-12-01DOI: 10.1109/BIBM.2016.7822603
S. N. Chandrasekaran, Jun Huan
The paradigm of drug discovery has moved from finding new drugs that exhibit therapeutic properties for a disease to reusing existing approved drugs for a newer disease. The association between a drug and a disease involves a complex network of targets and pathways. In order to provide new insights, there has been a constant need for sophisticated tools that have the potential to discover new associations from the underlying drugs-disease interactions. In addition to computational tools, there has been an explosion of data available in terms of drugs, disease and their activity profiles. On one hand, researchers have been using existing machine learning tools that have shown great promise in predicting associations but on the other hand there has been a void in exploiting advance machine learning frameworks to handle this kind of data integration. In this paper, we propose a learning framework called weighted multi-view learning that is a variant of the Multi-view learning framework in which the views are assumed to contribute equally to the prediction whereas our method learns a weight for each view since we hypothesize that certain views might have better prediction capability than others.
{"title":"Weighted multiview learning for predicting drug-disease associations","authors":"S. N. Chandrasekaran, Jun Huan","doi":"10.1109/BIBM.2016.7822603","DOIUrl":"https://doi.org/10.1109/BIBM.2016.7822603","url":null,"abstract":"The paradigm of drug discovery has moved from finding new drugs that exhibit therapeutic properties for a disease to reusing existing approved drugs for a newer disease. The association between a drug and a disease involves a complex network of targets and pathways. In order to provide new insights, there has been a constant need for sophisticated tools that have the potential to discover new associations from the underlying drugs-disease interactions. In addition to computational tools, there has been an explosion of data available in terms of drugs, disease and their activity profiles. On one hand, researchers have been using existing machine learning tools that have shown great promise in predicting associations but on the other hand there has been a void in exploiting advance machine learning frameworks to handle this kind of data integration. In this paper, we propose a learning framework called weighted multi-view learning that is a variant of the Multi-view learning framework in which the views are assumed to contribute equally to the prediction whereas our method learns a weight for each view since we hypothesize that certain views might have better prediction capability than others.","PeriodicalId":345384,"journal":{"name":"2016 IEEE International Conference on Bioinformatics and Biomedicine (BIBM)","volume":"100 ","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2016-12-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"134190462","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2016-12-01DOI: 10.1109/BIBM.2016.7822657
Runzhi Li, Hongling Zhao, Yusong Lin, Andrew S. Maxwell, Chaoyang Zhang
A Multi-Label Problem Transformation Joint Classification (MLPTJC) method is developed to solve the multi-label classification problem for the health and disease risk prediction based on physical examination records. We adopt a multi-class classification problem transformation method to transform the multi-label classification problem to a multi-class classification problem. Then We propose a Joint Decomposition Subset Classifier method to reduce the infrequent label sets to deal with the imbalance learning problem. Based on MLPTJC, existing cost-sensitive multi-class classification algorithms can be used to train the prediction models. We conduct some experiments to evaluate the performance of the MLPTJC method. The Support Vector Machine (SVM) and Random Forest (RF) algorithms are used for multi-class classification learning. We use the 10-fold cross-validation and metrics such as Average Accuracy, Precision, Recall and F-measure to evaluate the performance. The real physical examination records were employed, which include 62 examination items and 110, 300 anonymous patients. 8 types of diseases were predicted. The experimental results show that the MLPTJC method has better performance in terms of accuracy.
{"title":"Multi-label classification for intelligent health risk prediction","authors":"Runzhi Li, Hongling Zhao, Yusong Lin, Andrew S. Maxwell, Chaoyang Zhang","doi":"10.1109/BIBM.2016.7822657","DOIUrl":"https://doi.org/10.1109/BIBM.2016.7822657","url":null,"abstract":"A Multi-Label Problem Transformation Joint Classification (MLPTJC) method is developed to solve the multi-label classification problem for the health and disease risk prediction based on physical examination records. We adopt a multi-class classification problem transformation method to transform the multi-label classification problem to a multi-class classification problem. Then We propose a Joint Decomposition Subset Classifier method to reduce the infrequent label sets to deal with the imbalance learning problem. Based on MLPTJC, existing cost-sensitive multi-class classification algorithms can be used to train the prediction models. We conduct some experiments to evaluate the performance of the MLPTJC method. The Support Vector Machine (SVM) and Random Forest (RF) algorithms are used for multi-class classification learning. We use the 10-fold cross-validation and metrics such as Average Accuracy, Precision, Recall and F-measure to evaluate the performance. The real physical examination records were employed, which include 62 examination items and 110, 300 anonymous patients. 8 types of diseases were predicted. The experimental results show that the MLPTJC method has better performance in terms of accuracy.","PeriodicalId":345384,"journal":{"name":"2016 IEEE International Conference on Bioinformatics and Biomedicine (BIBM)","volume":"75 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2016-12-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"134600494","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
京公网安备 11010802042870号
京ICP备2023020795号-1
扫码关注我们
求助内容：
应助结果提醒方式：