Search PubMed⌕ Search

Biomedical subjects

Alexander J Hartemink

Publications and source records attributed to Alexander J Hartemink.

16 recordsLinked to original sources
PubMed2006

Computational inference of neural information flow networks.

Determining how information flows along anatomical brain pathways is a fundamental requirement for understanding how animals perceive their environments, learn, and behave. Attempts to reveal such neural information flow have been made using linear computational methods, but neural interactions are known to be nonlinear. Here, we demonstrate that a dynamic Bayesian network (DBN) inference algorithm we originally developed to infer nonlinear transcriptional regulatory networks from gene expression data collected with microarrays is also successful at inferring nonlinear neural information flow networks from electrophysiology data collected with microelectrode arrays. The inferred networks we recover from the songbird auditory pathway are correctly restricted to a subset of known anatomical paths, are consistent with timing of the system, and reveal both the importance of reciprocal feedback in auditory processing and greater information flow to higher-order auditory areas when birds hear natural as opposed to synthetic sounds. A linear method applied to the same data incorrectly produces networks with information flow to non-neural tissue and over paths known not to exist. To our knowledge, this study represents the first biologically validated demonstration of an algorithm to successfully infer neural information flow networks.

Action Potentials↗
PubMed2006

Informative priors based on transcription factor structural class improve de novo motif discovery.

MOTIVATION: An important problem in molecular biology is to identify the locations at which a transcription factor (TF) binds to DNA, given a set of DNA sequences believed to be bound by that TF. In previous work, we showed that information in the DNA sequence of a binding site is sufficient to predict the structural class of the TF that binds it. In particular, this suggests that we can predict which locations in any DNA sequence are more likely to be bound by certain classes of TFs than others. Here, we argue that traditional methods for de novo motif finding can be significantly improved by adopting an informative prior probability that a TF binding site occurs at each sequence location. To demonstrate the utility of such an approach, we present priority, a powerful new de novo motif finding algorithm. RESULTS: Using data from TRANSFAC, we train three classifiers to recognize binding sites of basic leucine zipper, forkhead, and basic helix loop helix TFs. These classifiers are used to equip priority with three class-specific priors, in addition to a default prior to handle TFs of other classes. We apply priority and a number of popular motif finding programs to sets of yeast intergenic regions that are reported by ChIP-chip to be bound by particular TFs. priority identifies motifs the other methods fail to identify, and correctly predicts the structural class of the TF recognizing the identified binding sites. AVAILABILITY: Supplementary material and code can be found at http://www.cs.duke.edu/~amink/.

Algorithms↗
PubMed2006

Finding diagnostic biomarkers in proteomic spectra.

In seeking to find diagnostic biomarkers in proteomic spectra, two significant problems arise. First, not only is there noise in the measured intensity at each m/z value, but there is also noise in the measured m/z value itself. Second, the potential for overfitting is severe: it is easy to find features in the spectra that accurately discriminate disease states but have no biological meaning. We address these problems by developing and testing a series of steps for pre-processing proteomic spectra and extracting putatively meaningful features before presentation to feature selection and classification algorithms. These steps include an HMM-based latent spectrum extraction algorithm for fusing the information from multiple replicate spectra obtained from a single tissue sample, a simple algorithm for baseline correction based on a segmented convex hull, a peak identification and quantification algorithm, and a peak registration algorithm to align peaks from multiple tissue samples into common peak registers. We apply these steps to MALDI spectral data collected from normal and tumor lung tissue samples, and then compare the performance of feature selection with FDR followed by classification with an SVM, versus joint feature selection and classification with Bayesian sparse multinomial logistic regression (SMLR). The SMLR approach outperformed FDR+SVM, but both were effective in achieving good diagnostic accuracy with a small number of features. Some of the selected features have previously been investigated as clinical markers for lung cancer diagnosis; some of the remaining features are excellent candidates for further research.

Algorithms↗
PubMed2005

Sequence features of DNA binding sites reveal structural class of associated transcription factor.

MOTIVATION: A key goal in molecular biology is to understand the mechanisms by which a cell regulates the transcription of its genes. One important aspect of this transcriptional regulation is the binding of transcription factors (TFs) to their specific cis-regulatory counterparts on the DNA. TFs recognize and bind their DNA counterparts according to the structure of their DNA-binding domains (e.g. zinc finger, leucine zipper, homeodomain). The structure of these domains can be used as a basis for grouping TFs into classes. Although the structure of DNA-binding domains varies widely across TFs generally, the TFs within a particular class bind to DNA in a similar fashion, suggesting the existence of class-specific features in the DNA sequences bound by each class of TFs. RESULTS: In this paper, we apply a sparse Bayesian learning algorithm to identify a small set of class-specific features in the DNA sequences bound by different classes of TFs; the algorithm simultaneously learns a true multi-class classifier that uses these features to predict the DNA-binding domain of the TF that recognizes a particular set of DNA sequences. We train our algorithm on the six largest classes in TRANSFAC, comprising a total of 587 TFs. We learn a six-class classifier for this training set that achieves 87% leave-one-out cross-validation accuracy. We also identify features within cis-regulatory sequences that are highly specific to each class of TF, which has significant implications for how TF binding sites should be modeled for the purpose of motif discovery.

Algorithms↗
PubMed2005

Genome-wide prediction of imprinted murine genes.

Imprinted genes are epigenetically modified genes whose expression is determined according to their parent of origin. They are involved in embryonic development, and imprinting dysregulation is linked to cancer, obesity, diabetes, and behavioral disorders such as autism and bipolar disease. Herein, we train a statistical model based on DNA sequence characteristics that not only identifies potentially imprinted genes, but also predicts the parental allele from which they are expressed. Of 23,788 annotated autosomal mouse genes, our model identifies 600 (2.5%) to be potentially imprinted, 64% of which are predicted to exhibit maternal expression. These predictions allowed for the identification of putative candidate genes for complex conditions where parent-of-origin effects are involved, including Alzheimer disease, autism, bipolar disorder, diabetes, male sexual orientation, obesity, and schizophrenia. We observe that the number, type, and relative orientation of repeated elements flanking a gene are particularly important in predicting whether a gene is imprinted.

Animals↗
PubMed2005

Sparse multinomial logistic regression: fast algorithms and generalization bounds.

Recently developed methods for learning sparse classifiers are among the state-of-the-art in supervised learning. These methods learn classifiers that incorporate weighted sums of basis functions with sparsity-promoting priors encouraging the weight estimates to be either significantly large or exactly zero. From a learning-theoretic perspective, these methods control the capacity of the learned classifier by minimizing the number of basis functions used, resulting in better generalization. This paper presents three contributions related to learning sparse classifiers. First, we introduce a true multiclass formulation based on multinomial logistic regression. Second, by combining a bound optimization approach with a component-wise update procedure, we derive fast exact algorithms for learning sparse multiclass classifiers that scale favorably in both the number of training samples and the feature dimensionality, making them applicable even to large data sets in high-dimensional feature spaces. To the best of our knowledge, these are the first algorithms to perform exact multinomial logistic regression with a sparsity-promoting prior. Third, we show how nontrivial generalization bounds can be derived for our classifier in the binary case. Experimental results on standard benchmark data sets attest to the accuracy, sparsity, and efficiency of the proposed methods.

Algorithms↗
PubMed2005

Informative structure priors: joint learning of dynamic regulatory networks from multiple types of data.

We present a method for jointly learning dynamic models of transcriptional regulatory networks from gene expression data and transcription factor binding location data. Models are automatically learned using dynamic Bayesian network inference algorithms; joint learning is accomplished by incorporating evidence from gene expression data through the likelihood, and from transcription factor binding location data through the prior. We propose a new informative structure prior with two advantages. First, the prior incorporates evidence from location data probabilistically, allowing it to be weighed against evidence from expression data. Second, the prior takes on a factorable form that is computationally efficient when learning dynamic regulatory networks. Results obtained from both simulated and experimental data from the yeast cell cycle demonstrate that this joint learning algorithm can recover dynamic regulatory networks from multiple types of data that are more accurate than those recovered from each type of data in isolation.

Bayes Theorem↗
PubMed2004

Theoretical and practical advances in genome halving.

MOTIVATION: Duplication of an organism's entire genome is a rare but spectacular event, enabling the rapid emergence of multiple new gene functions. Over time, the parallel linkage of duplicated genes across chromosomes may be disrupted by reciprocal translocations, while the intra-chromosomal order of genes may be shuffled by inversions and transpositions. Some duplicate genes may evolve unrecognizably or be deleted. As a consequence, the only detectable signature of an ancient duplication event in a modern genome may be the presence of various chromosomal segments containing parallel paralogous genes, with each segment appearing exactly twice in the genome. The problem of reconstructing the linkage structure of an ancestral genome before duplication is known as genome halving with unordered chromosomes. RESULTS: In this paper, we derive a new upper bound on the genome halving distance that is tighter than the best known, and a new lower bound that is almost always tighter than the best known. We also define the notion of genome halving diameter, and obtain both upper and lower bounds for it. Our tighter bounds on genome halving distance yield a new algorithm for reconstructing an ancestral duplicated genome. We create a software package GenomeHalving based on this new algorithm and test it on the yeast genome, identifying a sequence of translocations for halving the yeast genome that is shorter than previously conjectured possible.

Algorithms↗
PubMed2004

Advances to Bayesian network inference for generating causal networks from observational biological data.

MOTIVATION: Network inference algorithms are powerful computational tools for identifying putative causal interactions among variables from observational data. Bayesian network inference algorithms hold particular promise in that they can capture linear, non-linear, combinatorial, stochastic and other types of relationships among variables across multiple levels of biological organization. However, challenges remain when applying these algorithms to limited quantities of experimental data collected from biological systems. Here, we use a simulation approach to make advances in our dynamic Bayesian network (DBN) inference algorithm, especially in the context of limited quantities of biological data. RESULTS: We test a range of scoring metrics and search heuristics to find an effective algorithm configuration for evaluating our methodological advances. We also identify sampling intervals and levels of data discretization that allow the best recovery of the simulated networks. We develop a novel influence score for DBNs that attempts to estimate both the sign (activation or repression) and relative magnitude of interactions among variables. When faced with limited quantities of observational data, combining our influence score with moderate data interpolation reduces a significant portion of false positive interactions in the recovered networks. Together, our advances allow DBN inference algorithms to be more effective in recovering biological networks from experimentally collected data. AVAILABILITY: Source code and simulated data are available upon request. SUPPLEMENTARY INFORMATION: http://www.jarvislab.net/Bioinformatics/BNAdvances/

Algorithms↗
PubMed2004

Joint classifier and feature optimization for comprehensive cancer diagnosis using gene expression data.

Recent research has demonstrated quite convincingly that accurate cancer diagnosis can be achieved by constructing classifiers that are designed to compare the gene expression profile of a tissue of unknown cancer status to a database of stored expression profiles from tissues of known cancer status. This paper introduces the JCFO, a novel algorithm that uses a sparse Bayesian approach to jointly identify both the optimal nonlinear classifier for diagnosis and the optimal set of genes on which to base that diagnosis. We show that the diagnostic classification accuracy of the proposed algorithm is superior to a number of current state-of-the-art methods in a full leave-one-out cross-validation study of five widely used benchmark datasets. In addition to its superior classification accuracy, the algorithm is designed to automatically identify a small subset of genes (typically around twenty in our experiments) that are capable of providing complete discriminatory information for diagnosis. Focusing attention on a small subset of genes is useful not only because it produces a classifier with good generalization capacity, but also because this set of genes may provide insights into the mechanisms responsible for the disease itself. A number of the genes identified by the JCFO in our experiments are already in use as clinical markers for cancer diagnosis; some of the remaining genes may be excellent candidates for further clinical investigation. If it is possible to identify a small set of genes that is indeed capable of providing complete discrimination, inexpensive diagnostic assays might be widely deployable in clinical settings.

Algorithms↗
PubMed2004

A Bayesian approach to joint feature selection and classifier design.

This paper adopts a Bayesian approach to simultaneously learn both an optimal nonlinear classifier and a subset of predictor variables (or features) that are most relevant to the classification task. The approach uses heavy-tailed priors to promote sparsity in the utilization of both basis functions and features; these priors act as regularizers for the likelihood function that rewards good classification on the training data. We derive an expectation-maximization (EM) algorithm to efficiently compute a maximum a posteriori (MAP) point estimate of the various parameters. The algorithm is an extension of recent state-of-the-art sparse Bayesian classifiers, which in turn can be seen as Bayesian counterparts of support vector machines. Experimental comparisons using kernel classifiers demonstrate both parsimonious feature selection and excellent classification accuracy on a range of synthetic and benchmark data sets.

Algorithms↗
PubMed2003

Influence of network topology and data collection on network inference.

We recently developed an approach for testing the accuracy of network inference algorithms by applying them to biologically realistic simulations with known network topology. Here, we seek to determine the degree to which the network topology and data sampling regime influence the ability of our Bayesian network inference algorithm, NETWORKINFERENCE, to recover gene regulatory networks. NETWORKINFERENCE performed well at recovering feedback loops and multiple targets of a regulator with small amounts of data, but required more data to recover multiple regulators of a gene. When collecting the same number of data samples at different intervals from the system, the best recovery was produced by sampling intervals long enough such that sampling covered propagation of regulation through the network but not so long such that intervals missed internal dynamics. These results further elucidate the possibilities and limitations of network inference based on biological data.

Algorithms↗
PubMed2002

Evaluating functional network inference using simulations of complex biological systems.

MOTIVATION: Although many network inference algorithms have been presented in the bioinformatics literature, no suitable approach has been formulated for evaluating their effectiveness at recovering models of complex biological systems from limited data. To overcome this limitation, we propose an approach to evaluate network inference algorithms according to their ability to recover a complex functional network from biologically reasonable simulated data. RESULTS: We designed a simulator to generate data representing a complex biological system at multiple levels of organization: behaviour, neural anatomy, brain electrophysiology, and gene expression of songbirds. About 90% of the simulated variables are unregulated by other variables in the system and are included simply as distracters. We sampled the simulated data at intervals as one would sample from a biological system in practice, and then used the sampled data to evaluate the effectiveness of an algorithm we developed for functional network inference. We found that our algorithm is highly effective at recovering the functional network structure of the simulated system-including the irrelevance of unregulated variables-from sampled data alone. To assess the reproducibility of these results, we tested our inference algorithm on 50 separately simulated sets of data and it consistently recovered almost perfectly the complex functional network structure underlying the simulated data. To our knowledge, this is the first approach for evaluating the effectiveness of functional network inference algorithms at recovering models from limited data. Our simulation approach also enables researchers a priori to design experiments and data-collection protocols that are amenable to functional network inference.

Adaptation, Physiological↗
PubMed2002

Combining location and expression data for principled discovery of genetic regulatory network models.

We develop principled methods for the automatic induction (discovery) of genetic regulatory network models from multiple data sources and data modalities. Models of regulatory networks are represented as Bayesian networks, allowing the models to compactly and robustly capture probabilistic multivariate statistical dependencies between the various cellular factors in these networks. We build on previous Bayesian network validation results by extending the validation framework to the context of model induction, leveraging heuristic simulated annealing search algorithms and posterior model averaging. Using expression data in isolation yields results inconsistent with location data so we incorporate genomic location data to guide the model induction process. We combine these two data modalities by allowing location data to influence the model prior and expression data to influence the model likelihood. We demonstrate the utility of this approach by discovering genetic regulatory models of thirty-three variables involved in S. cerevisiae pheromone response. The models we automatically generate are consistent with the current understanding regarding this regulatory network, but also suggest new directions for future experimental investigation.

Bayes Theorem↗