Pub Date : 2012-10-04DOI: 10.1109/BIBMW.2012.6470216
Anurag Nagar, Michael Hahsler
This paper presents an alignment-free technique to efficiently discover similar regions in large sets of biological sequences using position sensitive p-mer frequency clustering. A set of sequences is broken down into segment and then a frequency distribution over all oligomers of size p (referred to as p-mers) is obtained to summarize each segment. These summaries are clustered while the order of segments in the set of sequences is preserved in a Markov-type model. Sequence segments within each cluster have very similar DNA/RNA patterns and form a so called quasi-alignment. This fact can be used for a variety of tasks such as species characterization and identification, phylogenetic analysis, functional analysis of sequences and, as in this paper, for discovering conserved regions. Our method is computationally more efficient than multiple sequences alignment since it can apply modern data stream clustering algorithms which run in time linear in the number of segments and thus can help discover highly similar regions across a large number of sequences efficiently. In this paper, we apply the approach to efficiently discover and visualize conserved regions in 16S rRNA.
{"title":"A novel quasi-alignment-based method for discovering conserved regions in genetic sequences","authors":"Anurag Nagar, Michael Hahsler","doi":"10.1109/BIBMW.2012.6470216","DOIUrl":"https://doi.org/10.1109/BIBMW.2012.6470216","url":null,"abstract":"This paper presents an alignment-free technique to efficiently discover similar regions in large sets of biological sequences using position sensitive p-mer frequency clustering. A set of sequences is broken down into segment and then a frequency distribution over all oligomers of size p (referred to as p-mers) is obtained to summarize each segment. These summaries are clustered while the order of segments in the set of sequences is preserved in a Markov-type model. Sequence segments within each cluster have very similar DNA/RNA patterns and form a so called quasi-alignment. This fact can be used for a variety of tasks such as species characterization and identification, phylogenetic analysis, functional analysis of sequences and, as in this paper, for discovering conserved regions. Our method is computationally more efficient than multiple sequences alignment since it can apply modern data stream clustering algorithms which run in time linear in the number of segments and thus can help discover highly similar regions across a large number of sequences efficiently. In this paper, we apply the approach to efficiently discover and visualize conserved regions in 16S rRNA.","PeriodicalId":6392,"journal":{"name":"2012 IEEE International Conference on Bioinformatics and Biomedicine Workshops","volume":"75 1","pages":"662-669"},"PeriodicalIF":0.0,"publicationDate":"2012-10-04","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"86114842","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 : 2012-10-04DOI: 10.1109/BIBM.2012.6392732
Mingyu Shao, Yi Yang, J. Guan, Shuigeng Zhou
This paper presents a comprehensive comparison study on the performances of major existing models over two PPI datasets, by comparing the global and local statistical properties of the original PPI networks and the model-reproduced ones. Our experimental results show that the DD model has best fitting ability while iSite model and STICKY model also fit well with the PPI datasets over most statistical properties.
{"title":"A comparison study on protein-protein interaction network models","authors":"Mingyu Shao, Yi Yang, J. Guan, Shuigeng Zhou","doi":"10.1109/BIBM.2012.6392732","DOIUrl":"https://doi.org/10.1109/BIBM.2012.6392732","url":null,"abstract":"This paper presents a comprehensive comparison study on the performances of major existing models over two PPI datasets, by comparing the global and local statistical properties of the original PPI networks and the model-reproduced ones. Our experimental results show that the DD model has best fitting ability while iSite model and STICKY model also fit well with the PPI datasets over most statistical properties.","PeriodicalId":6392,"journal":{"name":"2012 IEEE International Conference on Bioinformatics and Biomedicine Workshops","volume":"78 1","pages":"1-4"},"PeriodicalIF":0.0,"publicationDate":"2012-10-04","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"83931104","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 : 2012-10-04DOI: 10.1109/BIBMW.2012.6470374
Lei Zhang, W. Mao, Yin Li, G. Su, Xusheng Liu
To provide inspiration for developing new therapies of diabetic nephropathy, our study combed ancient formulas for diabetic nephropathy recorded in China from Tang dynasty to Qing dynasty, and discussed their application in modern times. A total of 87 ancient formulas for diabetic nephropathy were collected. Six-Ingredient Rehmannia Pill and Supplemented Kidney Qi Pill were recorded at higher frequency. Radix Astragali liqid, Xuan Bu Pill, and Ass Hide Glue Decoction were the earliest formulas for diabetic nephropathy recorded in Tang dynasty. Ginseng Powder was recorded by doctors in 4 dynasties, with the highest dynastic repeated frequency. Only formulas with high recorded frequency, such as Six-Ingredient Rehmannia Pill, Supplemented Kidney Qi Pill, Poria Pill, Four Ingredients Decoction, Pilose Antler Pill, were applied or studied in modern times keeping original medicinal combination. Most ancient formulas have not been made good use, and just sovereign medicinals in them were applied. In order to have better guidance from valuable experiences in ancient China, we should pay more attention to apply and study ancient formulas with original combination, not only the single Chinese medicinal.
{"title":"Gifts from Chinese Medicine for diabetic nephropathy: Ancient formulas in modern times","authors":"Lei Zhang, W. Mao, Yin Li, G. Su, Xusheng Liu","doi":"10.1109/BIBMW.2012.6470374","DOIUrl":"https://doi.org/10.1109/BIBMW.2012.6470374","url":null,"abstract":"To provide inspiration for developing new therapies of diabetic nephropathy, our study combed ancient formulas for diabetic nephropathy recorded in China from Tang dynasty to Qing dynasty, and discussed their application in modern times. A total of 87 ancient formulas for diabetic nephropathy were collected. Six-Ingredient Rehmannia Pill and Supplemented Kidney Qi Pill were recorded at higher frequency. Radix Astragali liqid, Xuan Bu Pill, and Ass Hide Glue Decoction were the earliest formulas for diabetic nephropathy recorded in Tang dynasty. Ginseng Powder was recorded by doctors in 4 dynasties, with the highest dynastic repeated frequency. Only formulas with high recorded frequency, such as Six-Ingredient Rehmannia Pill, Supplemented Kidney Qi Pill, Poria Pill, Four Ingredients Decoction, Pilose Antler Pill, were applied or studied in modern times keeping original medicinal combination. Most ancient formulas have not been made good use, and just sovereign medicinals in them were applied. In order to have better guidance from valuable experiences in ancient China, we should pay more attention to apply and study ancient formulas with original combination, not only the single Chinese medicinal.","PeriodicalId":6392,"journal":{"name":"2012 IEEE International Conference on Bioinformatics and Biomedicine Workshops","volume":"112 1","pages":"507-510"},"PeriodicalIF":0.0,"publicationDate":"2012-10-04","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"83938434","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 : 2012-10-04DOI: 10.1109/BIBM.2012.6392639
Michael Farnum, V. Lobanov, Michael Brennan, D. Agrafiotis, J. Kolpak, J. Ciervo, L. Alquier
Monitoring ongoing clinical trials is a crucial activity for sponsors mandated by the FDA. Some of the important goals are the early detection of safety issues, assurance of proper trial conduct at remote sites, and tracking efficiency and cost. Generating a comprehensive picture of the current state of a trial involves integrating a variety of data sources, including case report forms, laboratory results from contract labs, and safety reports. Since the volume of data is large and is updated repeatedly during the trial, tools to assist in quickly and thoroughly interrogating the data are greatly needed. Of particular importance is the ability to see both aggregate views containing multiple patients as well as the capacity to drill down to individual patients' data points. Here, we present Clinical Case, a tool that has been developed within the Janssen organization and is used for this purpose. Clinical Case provides the ability to quickly integrate SDTM datasets, define both standard and ad hoc data views, and provide regular data updates. Both standard and user-configured views can be persisted, which can be composed of a variety of interactive graphics, including typical visualizations, such as box plots, scatter plots, line charts, tree maps, heat maps, etc., as well as visualizations specifically designed for clinical data, such as the Hy's Law plot, patient timelines plot, and integrated subject listings. The user can define subsets of patients for filtering and highlighting data and use these to compare data across multiple domains. Integrated into the system is the ability to manually annotate patients and data, as well as communicate with trial administrators.
{"title":"Clinical Case: Enhancing medical monitoring with visualization and analytics","authors":"Michael Farnum, V. Lobanov, Michael Brennan, D. Agrafiotis, J. Kolpak, J. Ciervo, L. Alquier","doi":"10.1109/BIBM.2012.6392639","DOIUrl":"https://doi.org/10.1109/BIBM.2012.6392639","url":null,"abstract":"Monitoring ongoing clinical trials is a crucial activity for sponsors mandated by the FDA. Some of the important goals are the early detection of safety issues, assurance of proper trial conduct at remote sites, and tracking efficiency and cost. Generating a comprehensive picture of the current state of a trial involves integrating a variety of data sources, including case report forms, laboratory results from contract labs, and safety reports. Since the volume of data is large and is updated repeatedly during the trial, tools to assist in quickly and thoroughly interrogating the data are greatly needed. Of particular importance is the ability to see both aggregate views containing multiple patients as well as the capacity to drill down to individual patients' data points. Here, we present Clinical Case, a tool that has been developed within the Janssen organization and is used for this purpose. Clinical Case provides the ability to quickly integrate SDTM datasets, define both standard and ad hoc data views, and provide regular data updates. Both standard and user-configured views can be persisted, which can be composed of a variety of interactive graphics, including typical visualizations, such as box plots, scatter plots, line charts, tree maps, heat maps, etc., as well as visualizations specifically designed for clinical data, such as the Hy's Law plot, patient timelines plot, and integrated subject listings. The user can define subsets of patients for filtering and highlighting data and use these to compare data across multiple domains. Integrated into the system is the ability to manually annotate patients and data, as well as communicate with trial administrators.","PeriodicalId":6392,"journal":{"name":"2012 IEEE International Conference on Bioinformatics and Biomedicine Workshops","volume":"23 1","pages":"1-1"},"PeriodicalIF":0.0,"publicationDate":"2012-10-04","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"82810602","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 : 2012-10-04DOI: 10.1109/BIBMW.2012.6470255
Sushmita Paul, P. Maji
The microRNAs, also known as miRNAs are, the class of small non-coding RNAs that repress the expression of a gene post-transcriptionally. In effect, they regulate expression of a gene or protein. It has been observed that they play an important role in various cellular processes and thus help in carrying out normal functioning of a cell. However, dysregulation of miRNAs is found to be a major cause of a disease. Various studies have also shown the role of miRNAs in cancer and utility of miRNAs for the diagnosis of cancer and other diseases. A large number of works have been conducted to identify differentially expressed miRNAs as unlike with mRNA expression, a modest number of miRNAs might be sufficient to classify human cancers. In this regard, this paper presents a rough set based feature selection algorithm to select miRNAs from expression data that can classify tissue samples into their respective category with minimal error rate. It selects a set of miRNAs by maximizing both the relevance and significance of miRNAs. The effectiveness of the rough set based algorithm, along with a comparison with other related algorithms, is demonstrated on three miRNA microarray expression data sets using the B.632+ bootstrap error rate of support vector machine.
{"title":"Rough sets and support vector machine for selecting differentially expressed miRNAs","authors":"Sushmita Paul, P. Maji","doi":"10.1109/BIBMW.2012.6470255","DOIUrl":"https://doi.org/10.1109/BIBMW.2012.6470255","url":null,"abstract":"The microRNAs, also known as miRNAs are, the class of small non-coding RNAs that repress the expression of a gene post-transcriptionally. In effect, they regulate expression of a gene or protein. It has been observed that they play an important role in various cellular processes and thus help in carrying out normal functioning of a cell. However, dysregulation of miRNAs is found to be a major cause of a disease. Various studies have also shown the role of miRNAs in cancer and utility of miRNAs for the diagnosis of cancer and other diseases. A large number of works have been conducted to identify differentially expressed miRNAs as unlike with mRNA expression, a modest number of miRNAs might be sufficient to classify human cancers. In this regard, this paper presents a rough set based feature selection algorithm to select miRNAs from expression data that can classify tissue samples into their respective category with minimal error rate. It selects a set of miRNAs by maximizing both the relevance and significance of miRNAs. The effectiveness of the rough set based algorithm, along with a comparison with other related algorithms, is demonstrated on three miRNA microarray expression data sets using the B.632+ bootstrap error rate of support vector machine.","PeriodicalId":6392,"journal":{"name":"2012 IEEE International Conference on Bioinformatics and Biomedicine Workshops","volume":"734 ","pages":"864-871"},"PeriodicalIF":0.0,"publicationDate":"2012-10-04","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"91549750","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 : 2012-10-04DOI: 10.1109/BIBMW.2012.6470370
A. Dhanik, J. McMurray, L. Kavraki
It is well known that computer-aided docking of large ligands, with many rotatable bonds, is extremely difficult. AutoDock is a widely used docking program that can dock small ligands, with upto 5 or 6 rotatable bonds, accurately and quickly. Docking of larger ligands, however, is not very accurate and is computationally expensive. In this paper we present an AutoDock-based incremental docking protocol which docks a large ligand to its target protein in increments. A fragment of the large ligand is first chosen and then docked. Best docked conformations are incrementally grown and docked again, and this process is repeated until all the atoms of the ligand are docked. Each docking operation is performed using AutoDock. However, in each docking operation only a small number of rotatable bonds are allowed to rotate. We did a systematic docking study on a dataset of 73 protein-ligand complexes derived from the core set of PDBbind database. The number of rotatable bonds in the ligands vary from 7 to 30. Docking experiments were done to evaluate the docking performance of the incremental protocol in comparison to AutoDock's standard protocol. Results from the study show that, on average over the dataset, docking of large ligands using our incremental protocol is 23-fold computationally faster than docking using AutoDock's standard protocol and also has comparable or better accuracy. We propose that, for docking large ligands, our incremental protocol can be used as an alternative to AutoDock's standard protocol.
{"title":"Auto dock-based incremental docking protocol to improve docking of large ligands","authors":"A. Dhanik, J. McMurray, L. Kavraki","doi":"10.1109/BIBMW.2012.6470370","DOIUrl":"https://doi.org/10.1109/BIBMW.2012.6470370","url":null,"abstract":"It is well known that computer-aided docking of large ligands, with many rotatable bonds, is extremely difficult. AutoDock is a widely used docking program that can dock small ligands, with upto 5 or 6 rotatable bonds, accurately and quickly. Docking of larger ligands, however, is not very accurate and is computationally expensive. In this paper we present an AutoDock-based incremental docking protocol which docks a large ligand to its target protein in increments. A fragment of the large ligand is first chosen and then docked. Best docked conformations are incrementally grown and docked again, and this process is repeated until all the atoms of the ligand are docked. Each docking operation is performed using AutoDock. However, in each docking operation only a small number of rotatable bonds are allowed to rotate. We did a systematic docking study on a dataset of 73 protein-ligand complexes derived from the core set of PDBbind database. The number of rotatable bonds in the ligands vary from 7 to 30. Docking experiments were done to evaluate the docking performance of the incremental protocol in comparison to AutoDock's standard protocol. Results from the study show that, on average over the dataset, docking of large ligands using our incremental protocol is 23-fold computationally faster than docking using AutoDock's standard protocol and also has comparable or better accuracy. We propose that, for docking large ligands, our incremental protocol can be used as an alternative to AutoDock's standard protocol.","PeriodicalId":6392,"journal":{"name":"2012 IEEE International Conference on Bioinformatics and Biomedicine Workshops","volume":"91 1","pages":"48-55"},"PeriodicalIF":0.0,"publicationDate":"2012-10-04","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"89149769","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 : 2012-10-04DOI: 10.1109/BIBM.2012.6392693
Chengwei Lei, Jianhua Ruan
Recent advances in high-throughput technology have dramatically increased the quantity of available protein-protein interaction (PPI) data and stimulated the development of many methods for predicting protein complexes, which are important in understanding the functional organization of protein-protein interaction networks in different biological processes. However, automated protein complex prediction from PPI data alone is significantly hindered by the high level of noise, sparseness, and highly skewed degree distribution of PPI networks. Here we present a novel network topology-based algorithm to remove spurious interactions and recover missing ones by computational predictions, and to increase the accuracy of protein complex prediction by reducing the impact of hub nodes. The key idea of our algorithm is that two proteins sharing some high-order topological similarities, which are measured by a novel random walk-based procedure, are likely interacting with each other and may belong to the same protein complex. Applying our algorithm to a yeast protein-protein interaction network, we found that the interactions in the reconstructed PPI network have more significant biological relevance than the original network, assessed by multiple types of information, including gene ontology, gene expression, essentiality, conservation between species, and known protein complexes. Comparison with several existing methods show that the network reconstructed by our method has the highest quality. Finally, using two independent graph clustering algorithms, we found that the reconstructed network has resulted in significantly improved prediction accuracy of protein complexes.
{"title":"A random walk based approach for improving protein-protein interaction network and protein complex prediction","authors":"Chengwei Lei, Jianhua Ruan","doi":"10.1109/BIBM.2012.6392693","DOIUrl":"https://doi.org/10.1109/BIBM.2012.6392693","url":null,"abstract":"Recent advances in high-throughput technology have dramatically increased the quantity of available protein-protein interaction (PPI) data and stimulated the development of many methods for predicting protein complexes, which are important in understanding the functional organization of protein-protein interaction networks in different biological processes. However, automated protein complex prediction from PPI data alone is significantly hindered by the high level of noise, sparseness, and highly skewed degree distribution of PPI networks. Here we present a novel network topology-based algorithm to remove spurious interactions and recover missing ones by computational predictions, and to increase the accuracy of protein complex prediction by reducing the impact of hub nodes. The key idea of our algorithm is that two proteins sharing some high-order topological similarities, which are measured by a novel random walk-based procedure, are likely interacting with each other and may belong to the same protein complex. Applying our algorithm to a yeast protein-protein interaction network, we found that the interactions in the reconstructed PPI network have more significant biological relevance than the original network, assessed by multiple types of information, including gene ontology, gene expression, essentiality, conservation between species, and known protein complexes. Comparison with several existing methods show that the network reconstructed by our method has the highest quality. Finally, using two independent graph clustering algorithms, we found that the reconstructed network has resulted in significantly improved prediction accuracy of protein complexes.","PeriodicalId":6392,"journal":{"name":"2012 IEEE International Conference on Bioinformatics and Biomedicine Workshops","volume":"1 1","pages":"1-6"},"PeriodicalIF":0.0,"publicationDate":"2012-10-04","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"83146804","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 : 2012-10-04DOI: 10.1109/BIBM.2012.6392645
Dan He
In this work, we model how biomedicai topics influence one another, given they are organized in a topic hierarchy, MeSH, in which the edges capture a parent-child/subsumption relationship among topics. This information enables studying influence of topics from a semantic perspective, which might be very important in analyzing topic evolution and is missing from the current literature. We first define a burst-based action for topics, which models upward momentum in popularity (or "elevated occurrences" of the topics), and use it to define two types of influence: accumulation influence and propagation influence. We then propose a model of influence between topics, and develop an efficient algorithm (TIPS) to identify influential topics. Experiments show that our model is successful at identifying influential topics and the algorithm is very efficient.
{"title":"Modeling semantic influence for biomedicai research topics using MeSH hierarchy","authors":"Dan He","doi":"10.1109/BIBM.2012.6392645","DOIUrl":"https://doi.org/10.1109/BIBM.2012.6392645","url":null,"abstract":"In this work, we model how biomedicai topics influence one another, given they are organized in a topic hierarchy, MeSH, in which the edges capture a parent-child/subsumption relationship among topics. This information enables studying influence of topics from a semantic perspective, which might be very important in analyzing topic evolution and is missing from the current literature. We first define a burst-based action for topics, which models upward momentum in popularity (or \"elevated occurrences\" of the topics), and use it to define two types of influence: accumulation influence and propagation influence. We then propose a model of influence between topics, and develop an efficient algorithm (TIPS) to identify influential topics. Experiments show that our model is successful at identifying influential topics and the algorithm is very efficient.","PeriodicalId":6392,"journal":{"name":"2012 IEEE International Conference on Bioinformatics and Biomedicine Workshops","volume":"9 1","pages":"1-6"},"PeriodicalIF":0.0,"publicationDate":"2012-10-04","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"87358502","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 : 2012-10-04DOI: 10.1109/BIBMW.2012.6470360
Yefeng Cai, Yue Zhang, Zhao-hui Liang
Comparative effectiveness research (CER) is a new clinical study model featured by its strategic framework consists of four categories and three themes. The core strategy of CER is to conduct observational longitude research supported by electronic registry and large database based on real world practice. Since CER studies do not uses a classic randomized control trial (RCT) design, the well-developed data analytic methods for RCTs are challenged. The data groups which are not acquired from the same time point, or have significant difference at the baseline are unable to be compared by the classic differential statistical methods, or the outcome will be without robust statistical support. In this paper, we described the characteristics of the Zheng studies of Chinese medicine. Then some data analytic methods based on machine learning are introduced as potential solutions for the data processing in the CER research of Chinese medicine. Finally, a new strategic framework is introduced to establish the CER methodology for Chinese medicine.
{"title":"Application of data mining to Zheng studies of Chinese medicine based on CER","authors":"Yefeng Cai, Yue Zhang, Zhao-hui Liang","doi":"10.1109/BIBMW.2012.6470360","DOIUrl":"https://doi.org/10.1109/BIBMW.2012.6470360","url":null,"abstract":"Comparative effectiveness research (CER) is a new clinical study model featured by its strategic framework consists of four categories and three themes. The core strategy of CER is to conduct observational longitude research supported by electronic registry and large database based on real world practice. Since CER studies do not uses a classic randomized control trial (RCT) design, the well-developed data analytic methods for RCTs are challenged. The data groups which are not acquired from the same time point, or have significant difference at the baseline are unable to be compared by the classic differential statistical methods, or the outcome will be without robust statistical support. In this paper, we described the characteristics of the Zheng studies of Chinese medicine. Then some data analytic methods based on machine learning are introduced as potential solutions for the data processing in the CER research of Chinese medicine. Finally, a new strategic framework is introduced to establish the CER methodology for Chinese medicine.","PeriodicalId":6392,"journal":{"name":"2012 IEEE International Conference on Bioinformatics and Biomedicine Workshops","volume":"23 1","pages":"448-451"},"PeriodicalIF":0.0,"publicationDate":"2012-10-04","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"82028024","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 : 2012-10-04DOI: 10.1109/BIBM.2012.6392656
E. Fertig, Alexander V. Favorov, M. Ochs
Numerous methodologies, assays, and databases presently provide candidate targets of transcription factors (TFs). However, TFs rarely regulate their targets universally. The context of activation of a TF can change the transcriptional response of targets. Direct multiple regulation typical to mammalian genes complicates direct inference of TF targets from gene expression data. We present a novel statistic that infers context-specific TF regulation based upon the CoGAPS algorithm, which infers overlapping gene expression patterns resulting from coregulation. Numerical experiments with simulated data showed that this statistic correctly inferred targets that are common to multiple TFs, except in cases where the signal from a TF is negligible relative to noise level and signal from other TFs. The statistic is robust to moderate levels of error in the simulated gene sets, identifying fewer false positives than false negatives. Significantly, the regulatory statistic refines the number of transcription factor targets relevant to cell signaling in gastrointestinal stromal tumors (GIST) to genes consistent with the phosphorylation patterns of TFs identified in previous studies. As formulated, the proposed regulatory statistic has wide applicability to inferring set membership in integrated datasets. This statistic could be naturally extended to account for prior probabilities of set membership or to add candidate gene targets.
{"title":"Identifying context-specific transcription factor targets from prior knowledge and gene expression data","authors":"E. Fertig, Alexander V. Favorov, M. Ochs","doi":"10.1109/BIBM.2012.6392656","DOIUrl":"https://doi.org/10.1109/BIBM.2012.6392656","url":null,"abstract":"Numerous methodologies, assays, and databases presently provide candidate targets of transcription factors (TFs). However, TFs rarely regulate their targets universally. The context of activation of a TF can change the transcriptional response of targets. Direct multiple regulation typical to mammalian genes complicates direct inference of TF targets from gene expression data. We present a novel statistic that infers context-specific TF regulation based upon the CoGAPS algorithm, which infers overlapping gene expression patterns resulting from coregulation. Numerical experiments with simulated data showed that this statistic correctly inferred targets that are common to multiple TFs, except in cases where the signal from a TF is negligible relative to noise level and signal from other TFs. The statistic is robust to moderate levels of error in the simulated gene sets, identifying fewer false positives than false negatives. Significantly, the regulatory statistic refines the number of transcription factor targets relevant to cell signaling in gastrointestinal stromal tumors (GIST) to genes consistent with the phosphorylation patterns of TFs identified in previous studies. As formulated, the proposed regulatory statistic has wide applicability to inferring set membership in integrated datasets. This statistic could be naturally extended to account for prior probabilities of set membership or to add candidate gene targets.","PeriodicalId":6392,"journal":{"name":"2012 IEEE International Conference on Bioinformatics and Biomedicine Workshops","volume":"14 1","pages":"1-6"},"PeriodicalIF":0.0,"publicationDate":"2012-10-04","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"91212120","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：