Alex Lewin

Haplotype and isoform specific expression estimation using multi-mapping RNA-seq reads

We present a novel pipeline and methodology for simultaneously estimating isoform expression and allelic imbalance in diploid organisms using RNA-seq data. We achieve this by modeling the expression of haplotype-specific isoforms. If unknown, the two parental isoform sequences can be individually reconstructed. A new statistical method, MMSEQ, deconvolves the mapping of reads to multiple transcripts (isoforms or haplotype-specific isoforms). Our software can take into account non-uniform read generation and works with paired-end reads.

A mixture model-based strategy for selecting sets of genes in multiclass response microarray experiments

MOTIVATION Multiclass response (MCR) experiments are those in which there are more than two classes to be compared. In these experiments, though the null hypothesis is simple, there are typically many patterns of gene expression changes across the different classes that led to complex alternatives. In this paper, we propose a new strategy for selecting genes in MCR that is based on a flexible mixture model for the marginal distribution of a modified F-statistic. Using this model, false positive and negative discovery rates can be estimated and combined to produce a rule for selecting a subset of genes. Moreover, the method proposed allows calculation of these rates for any predefined subset of genes. RESULTS We illustrate the performance our approach using simulated datasets and a real breast cancer microarray dataset. In this latter study, we investigate predefined subset of genes and point out interesting differences between three distinct biological pathways. AVAILABILITY http://www.bgx.org.uk/software.html

Cancer risks in populations living near landfill sites in Great Britain

Previous studies have raised concerns about possible excess risks of bladder, brain and hepatobiliary cancers and leukaemias near landfill sites. Several cancers have been implicated, but no consistent pattern has emerged. We present a large nationwide analysis of selected cancers near landfill sites in Great Britain. The base population comprised people living within 2 km of 9565 (from a total of 19196) landfill sites that were operational at some time from 1982 to 1997, with populations living more than 2 km from a landfill as reference. Risks of cancers at the above sites were computed with adjustment for age, sex, year of diagnosis, region and deprivation. National post-coded registers provided a total of 341856640 person–years for the adult cancer analyses and 113631443 person–years for childhood leukaemia. There were 89786 cases of bladder cancer, 36802 cases of brain cancer, 21773 cases of hepatobiliary cancer, 37812 cases of adult leukaemia and 3973 cases of childhood leukaemia. In spite of the very large scale of this national study, we found no excess risks of cancers of the bladder and brain, hepatobiliary cancer or leukaemia, in populations living within 2 km of landfill sites. The results were similar if the analysis were restricted to landfill sites licensed to carry special (hazardous) waste. Our results do not support suggestions of excess risks of cancer associated with landfill sites reported in other studies.

Grouping Gene Ontology terms to improve the assessment of gene set enrichment in microarray data.

BackgroundGene Ontology (GO) terms are often used to assess the results of microarray experiments. The most common way to do this is to perform Fishers exact tests to find GO terms which are over-represented amongst the genes declared to be differentially expressed in the analysis of the microarray experiment. However, due to the high degree of dependence between GO terms, statistical testing is conservative, and interpretation is difficult.ResultsWe propose testing groups of GO terms rather than individual terms, to increase statistical power, reduce dependence between tests and improve the interpretation of results. We use the publicly available package POSOC to group the terms. Our method finds groups of GO terms significantly over-represented amongst differentially expressed genes which are not found by Fishers tests on individual GO terms.ConclusionGrouping Gene Ontology terms improves the interpretation of gene set enrichment for microarray data.

Optimal Whitening and Decorrelation

ABSTRACT Whitening, or sphering, is a common preprocessing step in statistical analysis to transform random variables to orthogonality. However, due to rotational freedom there are infinitely many possible whitening procedures. Consequently, there is a diverse range of sphering methods in use, for example, based on principal component analysis (PCA), Cholesky matrix decomposition, and zero-phase component analysis (ZCA), among others. Here, we provide an overview of the underlying theory and discuss five natural whitening procedures. Subsequently, we demonstrate that investigating the cross-covariance and the cross-correlation matrix between sphered and original variables allows to break the rotational invariance and to identify optimal whitening transformations. As a result we recommend two particular approaches: ZCA-cor whitening to produce sphered variables that are maximally similar to the original variables, and PCA-cor whitening to obtain sphered variables that maximally compress the original variables.

MMBGX: a method for estimating expression at the isoform level and detecting differential splicing using whole-transcript Affymetrix arrays

Affymetrix has recently developed whole-transcript GeneChips—‘Gene’ and ‘Exon’ arrays—which interrogate exons along the length of each gene. Although each probe on these arrays is intended to hybridize perfectly to only one transcriptional target, many probes match multiple transcripts located in different parts of the genome or alternative isoforms of the same gene. Existing statistical methods for estimating expression do not take this into account and are thus prone to producing inflated estimates. We propose a method, Multi-Mapping Bayesian Gene eXpression (MMBGX), which disaggregates the signal at ‘multi-match’ probes. When applied to Gene arrays, MMBGX removes the upward bias of gene-level expression estimates. When applied to Exon arrays, it can further disaggregate the signal between alternative transcripts of the same gene, providing expression estimates of individual splice variants. We demonstrate the performance of MMBGX on simulated data and a tissue mixture data set. We then show that MMBGX can estimate the expression of alternative isoforms within one experimental condition, confirming our results by RT-PCR. Finally, we show that our method for detecting differential splicing has a lower error rate than standard exon-level approaches on a previously validated colon cancer data set.

Sarcoidosis and tuberculosis cytokine profiles: indistinguishable in bronchoalveolar lavage but different in blood.

Background The clinical, radiological and pathological similarities between sarcoidosis and tuberculosis can make disease differentiation challenging. A complicating factor is that some cases of sarcoidosis may be initiated by mycobacteria. We hypothesised that immunological profiling might provide insight into a possible relationship between the diseases or allow us to distinguish between them. Methods We analysed bronchoalveolar lavage (BAL) fluid in sarcoidosis (n = 18), tuberculosis (n = 12) and healthy volunteers (n = 16). We further investigated serum samples in the same groups; sarcoidosis (n = 40), tuberculosis (n = 15) and healthy volunteers (n = 40). A cross-sectional analysis of multiple cytokine profiles was performed and data used to discriminate between samples. Results We found that BAL profiles were indistinguishable between both diseases and significantly different from healthy volunteers. In sera, tuberculosis patients had significantly lower levels of the Th2 cytokine interleukin-4 (IL-4) than those with sarcoidosis (p = 0.004). Additional serum differences allowed us to create a linear regression model for disease differentiation (within-sample accuracy 91%, cross-validation accuracy 73%). Conclusions These data warrant replication in independent cohorts to further develop and validate a serum cytokine signature that may be able to distinguish sarcoidosis from tuberculosis. Systemic Th2 cytokine differences between sarcoidosis and tuberculosis may also underly different disease outcomes to similar respiratory stimuli.

Testing for linkage and Hardy-Weinberg disequilibrium

This paper concerns several important points when testing for Hardy‐Weinberg equilibrium (HWE) and linkage disequilibrium (LD) in genetics. First, we challenge the necessity of using exclusively two‐sided tests for LD. Next, we show that the exact 2‐sided tests based on the most popular measures of LD are not equivalent, and neither are the standard statistical tests even though the 1‐sided tests are equivalent. We show how this results in different inference about LD for two data sets consisting of small groups of markers. Finally, we advocate the use of the conditional p‐value for both LD and HWE testing. An important advantage of this p‐value is that equivalent 1‐sided tests are transformed into equivalent 2‐sided tests.

MT-HESS: An efficient Bayesian approach for simultaneous association detection in OMICS datasets, with application to eQTL mapping in multiple tissues

Motivation: Analysing the joint association between a large set of responses and predictors is a fundamental statistical task in integrative genomics, exemplified by numerous expression Quantitative Trait Loci (eQTL) studies. Of particular interest are the so-called ‘hotspots’, important genetic variants that regulate the expression of many genes. Recently, attention has focussed on whether eQTLs are common to several tissues, cell-types or, more generally, conditions or whether they are specific to a particular condition. Results: We have implemented MT-HESS, a Bayesian hierarchical model that analyses the association between a large set of predictors, e.g. SNPs, and many responses, e.g. gene expression, in multiple tissues, cells or conditions. Our Bayesian sparse regression algorithm goes beyond ‘one-at-a-time’ association tests between SNPs and responses and uses a fully multivariate model search across all linear combinations of SNPs, coupled with a model of the correlation between condition/tissue-specific responses. In addition, we use a hierarchical structure to leverage shared information across different genes, thus improving the detection of hotspots. We show the increase of power resulting from our new approach in an extensive simulation study. Our analysis of two case studies highlights new hotspots that would remain undetected by standard approaches and shows how greater prediction power can be achieved when several tissues are jointly considered. Availability and implementation: C++ source code and documentation including compilation instructions are available under GNU licence at http://www.mrc-bsu.cam.ac.uk/software/. Contact: sylvia.richardson@mrc-bsu.cam.ac.uk or lb664@cam.ac.uk Supplementary information: Supplementary data are available at Bioinformatics online.

A new statistic for picking out Non-Gaussianity in the CMB

ABSTRA C T In this paper we propose a new statistic capable of detecting non-Gaussianity in the CMB. The statistic is defined in Fourier space, and therefore naturally separates angular scales. It consists of taking another Fourier transform, in angle, over the Fourier modes within a given ring of scales. Like other Fourier space statistics, our statistic outdoes more conventional methods when faced with combinations of Gaussian processes (be they noise or signal) and a nonGaussian signal which dominates only on some scales. However, unlike previous efforts along these lines, our statistic is successful in recognizing multiple non-Gaussian patterns in a single field. We discuss various applications, in which the Gaussian component may be noise or primordial signal, and the non-Gaussian component may be a cosmic string map, or some geometrical construction mimicking, say, small-scale dust maps.