Generic framework for high-dimensional fixed-effects ANOVA

In functional genomics it is more rule than exception that experimental designs are used to generate the data. The samples of the resulting data sets are thus organized according to this design and for each sample many biochemical compounds are measured, e.g. typically thousands of gene-expressions or hundreds of metabolites. This results in high-dimensional data sets with an underlying experimental design. Several methods have recently become available for analyzing such data while utilizing the underlying design. We review these methods by putting them in a unifying and general framework to facilitate understanding the (dis-)similarities between the methods. The biological question dictates which method to use and the framework allows for building new methods to accommodate a range of such biological questions. The framework is built on well known fixed-effect ANOVA models and subsequent dimension reduction. We present the framework both in matrix algebra as well as in more insightful geometrical terms. We show the workings of the different special cases of our framework with a real-life metabolomics example from nutritional research and a gene-expression example from the field of virology.     

A statistical framework for combining and interpreting proteomic datasets

MOTIVATION: To identify accurately protein function on a proteome-wide scale requires integrating data within and between high-throughput experiments. High-throughput proteomic datasets often have high rates of errors and thus yield incomplete and contradictory information. In this study, we develop a simple statistical framework using Bayes' law to interpret such data and combine information from different high-throughput experiments. In order to illustrate our approach we apply it to two protein complex purification datasets. RESULTS: Our approach shows how to use high-throughput data to calculate accurately the probability that two proteins are part of the same complex. Importantly, our approach does not need a reference set of verified protein interactions to determine false positive and false negative error rates of protein association. We also demonstrate how to combine information from two separate protein purification datasets into a combined dataset that has greater coverage and accuracy than either dataset alone. In addition, we also provide a technique for estimating the total number of proteins which can be detected using a particular experimental technique. AVAILABILITY: A suite of simple programs to accomplish some of the above tasks is available at www.unm.edu/~compbio/software/DatasetAssess     

A general framework of nonparametric feature selection in high-dimensional data

Nonparametric feature selection for high-dimensional data is an important and challenging problem in the fields of statistics and machine learning. Most of the existing methods for feature selection focus on parametric or additive models which may suffer from model misspecification. In this paper, we propose a new framework to perform nonparametric feature selection for both regression and classification problems. Under this framework, we learn prediction functions through empirical risk minimization over a reproducing kernel Hilbert space. The space is generated by a novel tensor product kernel, which depends on a set of parameters that determines the importance of the features. Computationally, we minimize the empirical risk with a penalty to estimate the prediction and kernel parameters simultaneously. The solution can be obtained by iteratively solving convex optimization problems. We study the theoretical property of the kernel feature space and prove the oracle selection property and Fisher consistency of our proposed method. Finally, we demonstrate the superior performance of our approach compared to existing methods via extensive simulation studies and applications to two real studies.     

A knowledge-driven agent-centred framework for data mining in EMG

In this paper, we present a multi-agent framework for data mining in electromyography. This application, based on a web interface, provides a set of functionalities allowing to manipulate 1000 medical cases and more than 25,000 neurological tests stored in a medical database. The aim is to extract medical information using data mining algorithms and to supply a knowledge base with pertinent information. The multi-agent platform gives the possibility to distribute the data management process between several autonomous entities. This framework provides a parallel and flexible data manipulation.     

Penalized generalized estimating equations for high-dimensional longitudinal data analysis

We consider the penalized generalized estimating equations (GEEs) for analyzing longitudinal data with high-dimensional covariates, which often arise in microarray experiments and large-scale health studies. Existing high-dimensional regression procedures often assume independent data and rely on the likelihood function. Construction of a feasible joint likelihood function for high-dimensional longitudinal data is challenging, particularly for correlated discrete outcome data. The penalized GEE procedure only requires specifying the first two marginal moments and a working correlation structure. We establish the asymptotic theory in a high-dimensional framework where the number of covariates p(n) increases as the number of clusters n increases, and p(n) can reach the same order as n. One important feature of the new procedure is that the consistency of model selection holds even if the working correlation structure is misspecified. We evaluate the performance of the proposed method using Monte Carlo simulations and demonstrate its application using a yeast cell-cycle gene expression data set.     

Efficient mining differential co-expression biclusters in microarray datasets

Background

Biclustering algorithm can find a number of co-expressed genes under a set of experimental conditions. Recently, differential co-expression bicluster mining has been used to infer the reasonable patterns in two microarray datasets, such as, normal and cancer cells.

Methods

In this paper, we propose an algorithm, DECluster, to mine Differential co-Expression biCluster in two discretized microarray datasets. Firstly, DECluster produces the differential co-expressed genes from each pair of samples in two microarray datasets, and constructs a differential weighted undirected sample–sample relational graph. Secondly, the differential biclusters are generated in the above differential weighted undirected sample–sample relational graph. In order to mine maximal differential co-expression biclusters efficiently, we design several pruning techniques for generating maximal biclusters without candidate maintenance.

Results

The experimental results show that our algorithm is more efficient than existing methods. The performance of DECluster is evaluated by empirical p-value and gene ontology, the results show that our algorithm can find more statistically significant and biological differential co-expression biclusters than other algorithms.

Conclusions

Our proposed algorithm can find more statistically significant and biological biclusters in two microarray datasets than the other two algorithms.     

A computational framework for gene regulatory network inference that combines multiple methods and datasets

Background  

Reverse engineering in systems biology entails inference of gene regulatory networks from observational data. This data typically include gene expression measurements of wild type and mutant cells in response to a given stimulus. It has been shown that when more than one type of experiment is used in the network inference process the accuracy is higher. Therefore the development of generally applicable and effective methodologies that embed multiple sources of information in a single computational framework is a worthwhile objective.     

On generalized latent factor modeling and inference for high-dimensional binomial data

We explore a hierarchical generalized latent factor model for discrete and bounded response variables and in particular, binomial responses. Specifically, we develop a novel two-step estimation procedure and the corresponding statistical inference that is computationally efficient and scalable for the high dimension in terms of both the number of subjects and the number of features per subject. We also establish the validity of the estimation procedure, particularly the asymptotic properties of the estimated effect size and the latent structure, as well as the estimated number of latent factors. The results are corroborated by a simulation study and for illustration, the proposed methodology is applied to analyze a dataset in a gene–environment association study.     

Comparison of sequence and structure-based datasets for nonredundant structural data mining

Structural data mining studies attempt to deduce general principles of protein structure from solved structures deposited in the protein data bank (PDB). The entire database is unsuitable for such studies because it is not representative of the ensemble of protein folds. Given that novel folds continue to be unearthed, some folds are currently unrepresented in the PDB while other folds are overrepresented. Overrepresentation can easily be avoided by filtering the dataset. PDB_SELECT is a well-used representative subset of the PDB that has been deduced by sequence comparison. Specifically, structures with sequences that exhibit a pairwise sequence identity above a threshold value are weeded from the dataset. Although length criteria for pairwise alignments have a structural basis, this automated method of pruning is essentially sequence-based and runs into problems in the twilight zone, possibly resulting in some folds being overrepresented. The value-added structure databases SCOP and CATH are also a potential source of a nonredundant dataset. Here we compare the sequence-derived dataset PDB_SELECT with the structural databases SCOP (Structural Classification Of Proteins) and CATH (Class-Architecture-Topology-Homology). We show that some folds remain overrepresented in the PDB_SELECT dataset while other folds are not represented at all. However, SCOP and CATH also have their own problems such as the labor-intensiveness of the update process and the problem of determining whether all folds are equally or sufficiently distant. We discuss areas where further work is required.     

A statistical framework for modeling gene expression using chromatin features and application to modENCODE datasets

We develop a statistical framework to study the relationship between chromatin features and gene expression. This can be used to predict gene expression of protein coding genes, as well as microRNAs. We demonstrate the prediction in a variety of contexts, focusing particularly on the modENCODE worm datasets. Moreover, our framework reveals the positional contribution around genes (upstream or downstream) of distinct chromatin features to the overall prediction of expression levels.     

A framework for analysing colour pattern geometry: adjacent colours

The analysis of colour pattern geometry is not as well advanced as the analysis of colour, although this reflects a lack of an analytical framework. The present study proposes an approach based on a consideration of which colours are adjacent to each other. Both vertebrate and invertebrate eyes do not take static images of the world but move across the field of view. As a consequence, the eye takes transects across the field of view responding to the colours and luminances within patches and to the colour and/or luminance transitions between patches. The framework and methods suggested here are based upon transects across colour patterns and make it possible to estimate colour pattern parameters that capture not only the relative areas of each patch class, but also the relative frequencies of colour/luminance transitions or adjacency. This allows tests of new hypotheses about colour patterns at the same time as including colour, pattern, and texture. Eleven groups of predictions are made with respect to the often conflicting needs of communication with conspecifics, avoiding predation, and finding food. New phenomena may be discovered as a result of these methods and predictions. For example, certain colour transitions may be used for species recognition even though the same colours are used by all species. © 2012 The Linnean Society of London, Biological Journal of the Linnean Society, 2012, ??, ??–??.     

Coev2Net: a computational framework for boosting confidence in high-throughput protein-protein interaction datasets

ABSTRACT: Improving the quality and coverage of the protein interactome is of tantamount importance for biomedical research, particularly given the various sources of uncertainty in high-throughput techniques. We introduce a structure-based framework, Coev2Net, for computing a single confidence score that addresses both false positive and false negative rates. Coev2Net is easily applied to thousands of binary protein interactions and has superior predictive performance over existing methods. We experimentally validate selected high-confidence predictions in the human MAPK network and show that predicted interfaces are enriched for cancer-related or damaging SNPs. Coev2Net can be downloaded at http://struct2net.csail.mit.edu/     

Quantitative mining of compositional heterogeneity in cryo-EM datasets of ribosome assembly intermediates

1. Download : Download high-res image (222KB)
2. Download : Download full-size image