Malaria represents one of the major worldwide challenges to public health. A recent breakthrough in study disease follows annotation genome malaria parasite Plasmodium falciparum and mosquito vector (an organism that spreads an infectious disease)Anopheles. Of particular interest is molecular biology underlying immune response system Anopheles, which actively fights against infection. This article reports a statistical analysis gene expression time profiles from mosquitoes have been infected...

Learning the network structure of a large graph is computationally demanding, and dynamically monitoring over time for any changes in threatens to be more challenging still. This paper presents two-stage method anomaly detection dynamic graphs: first stage uses simple, conjugate Bayesian models discrete counting processes track pairwise links all nodes assess normality behavior; second applies standard inference tools on greatly reduced subset potentially anomalous nodes. The utility...

Combining p-values from independent statistical tests is a popular approach to meta-analysis, particularly when the data underlying are either no longer available or difficult combine. A diverse range of p-value combination methods appear in literature, each with different properties. Yet all too often final choice used meta-analysis can arbitrary, as if effort has been expended building models that gave rise p-values. Birnbaum (1954) showed any reasonable combiner must be optimal against...

Abstract We introduce a procedure for generalized monotonic curve fitting that is based on Bayesian analysis of the isotonic regression model. Conventional fits monotonically increasing step functions to data. In our approach we treat number and location steps as random. For each level adopt conjugate prior sampling distribution data if was unconstrained. then propose use Markov chain Monte Carlo simulation draw samples from unconstrained model space retain only those which constraint holds....

We present a method for Bayesian model-based hierarchical coclustering of gene expression data and use it to study the temporal transcription responses an Anopheles gambiae cell line upon challenge with multiple microbial elicitors. The fits statistical regression models time series each experiment performs on genes by optimizing joint probability model, characterizing coregulation between experiments. compute model using two-stage Expectation-Maximization-type algorithm, first fixing...

The vast potential of the genomic insight offered by microarray technologies has led to their widespread use since they were introduced a decade ago. Application areas include gene function discovery, disease diagnosis, and inferring regulatory networks. Microarray experiments enable large‐scale, high‐throughput investigations activity have thus provided data analyst with distinctive, high‐dimensional field study. Many questions in this relate finding subgroups profiles which are very...

Abstract Spectral embedding of adjacency or Laplacian matrices undirected graphs is a common technique for representing network in lower dimensional latent space, with optimal theoretical guarantees. The can be used to estimate the community structure network, strong consistency results stochastic blockmodel framework. One main practical limitations standard algorithms detection from spectral embeddings that number communities and dimension must specified advance. In this article, novel...

In model-based clustering of complex data, a probability model, typically finite mixture forms the basis distance measure between any pair clusters. The idea was popularized by framework and accompanying software Fraley Raftery (2002). particular, agglomerative hierarchical is now frequently used approach for probabilistic grouping due to speed simplicity implementation. This article investigates deficiencies in clusterings proposed from this popular approach, presents review small...

When analysing multiple time series that may be subject to changepoints, it is sometimes possible specify a priori, by means of graph, which pairs are likely impacted simultaneous changepoints. This article proposes an informative prior for changepoints encodes the information contained in inducing changepoint model borrows strength across clusters connected detect weak signals synchronous The graphical further extended allow dependence between nearby but not necessarily neighbouring graph....

Connectivity patterns between nodes in a computer network can be interpreted and modelled as point processes where events process indicate connections being established for data to sent along that edge. A model of normal connectivity behaviour constructed each edge by identifying key user features such seasonality or self-exciting behaviour, since typically arise bursts at particular times day which may peculiar When monitoring real time, unusual activity against the normality could presence...

Monitoring computer network traffic for anomalous behaviour presents an important security challenge. Arrivals of new edges in a graph represent connections between client and server pair not previously observed, rare cases these might suggest the presence intruders or malicious implants. We propose Bayesian model anomaly detection method simultaneously characterising existing structure modelling likely edge formation. The is demonstrated on real authentication data successfully identifies...

Abstract Changepoint models typically assume the data within each segment are independent and identically distributed conditional on some parameters that change across segments. This construction may be inadequate when subject to local correlation patterns, often resulting in many more changepoints fitted than preferable. article proposes a Bayesian changepoint model relaxes assumption of exchangeability The proposed supposes m -dependent for unknown $$m \geqslant 0$$ <mml:math...

Graph link prediction is an important task in cyber-security: relationships between entities within a computer network, such as users interacting with computers, or system libraries and the corresponding processes that use them, can provide key insights into adversary behaviour. Poisson matrix factorisation (PMF) popular model for large networks, particularly useful its scalability. In this article, PMF extended to include scenarios are commonly encountered cyber-security applications....

A new class of models for dynamic networks is proposed, called mutually exciting point process graphs (MEG). MEG a scalable network-wide statistical model processes with dyadic marks, which can be used anomaly detection when assessing the significance future events, including previously unobserved connections between nodes. The combines to estimate dependencies events and latent space infer relationships intensity functions each network edge are characterised exclusively by node-specific...

Fission yeast Schizosaccharomyces pombe and budding Saccharomyces cerevisiae are among the original model organisms in study of cell-division cycle. Unlike yeast, no large-scale regulatory network has been constructed for fission yeast. It only partially characterized. As a result, important cascades have known or complete counterpart By integrating genome-wide data from multiple time course cell cycle microarray experiments we reconstructed gene network. Based on network, discovered...

Process monitoring and control requires the detection of structural changes in a data stream real time. This article introduces an efficient sequential Monte Carlo algorithm designed for learning unknown changepoints continuous The method is intuitively simple: new latest window are proposed by conditioning only on observed since most recent estimated changepoint, as these observations carry information about current state process. shows improved performance over art. Another advantage that...

Formulae display:?Mathematical formulae have been encoded as MathML and are displayed in this HTML version using MathJax order to improve their display. Uncheck the box turn off. This feature requires Javascript. Click on a formula zoom.

Abstract Periodic patterns can often be observed in real-world event time data, possibly mixed with non-periodic arrival times. For modelling purposes, it is necessary to correctly distinguish the two types of events. This task has particularly important implications computer network security; there, separating automated polling traffic and human-generated activity a for building realistic statistical models normal activity, which turn used anomaly detection. Since events commonly occur at...