Transcriptional, epigenetic and metabolic signatures in cardiometabolic syndrome defined by extreme phenotypes.

BACKGROUND: This work is aimed at improving the understanding of cardiometabolic syndrome pathophysiology and its relationship with thrombosis by generating a multi-omic disease signature. METHODS/RESULTS: We combined classic plasma biochemistry and plasma biomarkers with the transcriptional and epigenetic characterisation of cell types involved in thrombosis, obtained from two extreme phenotype groups (morbidly obese and lipodystrophy) and lean individuals to identify the molecular mechanisms at play, highlighting patterns of abnormal activation in innate immune phagocytic cells. Our analyses showed that extreme phenotype groups could be distinguished from lean individuals, and from each other, across all data layers. The characterisation of the same obese group, 6 months after bariatric surgery, revealed the loss of the abnormal activation of innate immune cells previously observed. However, rather than reverting to the gene expression landscape of lean individuals, this occurred via the establishment of novel gene expression landscapes. NETosis and its control mechanisms emerge amongst the pathways that show an improvement after surgical intervention. CONCLUSIONS: We showed that the morbidly obese and lipodystrophy groups, despite some differences, shared a common cardiometabolic syndrome signature. We also showed that this could be used to discriminate, amongst the normal population, those individuals with a higher likelihood of presenting with the disease, even when not displaying the classic features.

Tailored Bayes: a risk modeling framework under unequal misclassification costs.

Risk prediction models are a crucial tool in healthcare. Risk prediction models with a binary outcome (i.e., binary classification models) are often constructed using methodology which assumes the costs of different classification errors are equal. In many healthcare applications, this assumption is not valid, and the differences between misclassification costs can be quite large. For instance, in a diagnostic setting, the cost of misdiagnosing a person with a life-threatening disease as healthy may be larger than the cost of misdiagnosing a healthy person as a patient. In this article, we present Tailored Bayes (TB), a novel Bayesian inference framework which "tailors" model fitting to optimize predictive performance with respect to unbalanced misclassification costs. We use simulation studies to showcase when TB is expected to outperform standard Bayesian methods in the context of logistic regression. We then apply TB to three real-world applications, a cardiac surgery, a breast cancer prognostication task, and a breast cancer tumor classification task and demonstrate the improvement in predictive performance over standard methods.

The RNA landscape of the human placenta in health and disease.

The placenta is the interface between mother and fetus and inadequate function contributes to short and long-term ill-health. The placenta is absent from most large-scale RNA-Seq datasets. We therefore analyze long and small RNAs (~101 and 20 million reads per sample respectively) from 302 human placentas, including 94 cases of preeclampsia (PE) and 56 cases of fetal growth restriction (FGR). The placental transcriptome has the seventh lowest complexity of 50 human tissues: 271 genes account for 50% of all reads. We identify multiple circular RNAs and validate 6 of these by Sanger sequencing across the back-splice junction. Using large-scale mass spectrometry datasets, we find strong evidence of peptides produced by translation of two circular RNAs. We also identify novel piRNAs which are clustered on Chr1 and Chr14. PE and FGR are associated with multiple and overlapping differences in mRNA, lincRNA and circRNA but fewer consistent differences in small RNAs. Of the three protein coding genes differentially expressed in both PE and FGR, one encodes a secreted protein FSTL3 (follistatin-like 3). Elevated serum levels of FSTL3 in pregnant women are predictive of subsequent PE and FGR. To aid visualization of our placenta transcriptome data, we develop a web application ( https://www.obgyn.cam.ac.uk/placentome/ ).

Multiple kernel learning for integrative consensus clustering of omic datasets.

MOTIVATION: Diverse applications-particularly in tumour subtyping-have demonstrated the importance of integrative clustering techniques for combining information from multiple data sources. Cluster Of Clusters Analysis (COCA) is one such approach that has been widely applied in the context of tumour subtyping. However, the properties of COCA have never been systematically explored, and its robustness to the inclusion of noisy datasets is unclear. RESULTS: We rigorously benchmark COCA, and present Kernel Learning Integrative Clustering (KLIC) as an alternative strategy. KLIC frames the challenge of combining clustering structures as a multiple kernel learning problem, in which different datasets each provide a weighted contribution to the final clustering. This allows the contribution of noisy datasets to be down-weighted relative to more informative datasets. We compare the performances of KLIC and COCA in a variety of situations through simulation studies. We also present the output of KLIC and COCA in real data applications to cancer subtyping and transcriptional module discovery. AVAILABILITY AND IMPLEMENTATION: R packages klic and coca are available on the Comprehensive R Archive Network. SUPPLEMENTARY INFORMATION: Supplementary data are available at Bioinformatics online.

Fast approximate inference for variable selection in Dirichlet process mixtures, with an application to pan-cancer proteomics.

The Dirichlet Process (DP) mixture model has become a popular choice for model-based clustering, largely because it allows the number of clusters to be inferred. The sequential updating and greedy search (SUGS) algorithm (Wang & Dunson, 2011) was proposed as a fast method for performing approximate Bayesian inference in DP mixture models, by posing clustering as a Bayesian model selection (BMS) problem and avoiding the use of computationally costly Markov chain Monte Carlo methods. Here we consider how this approach may be extended to permit variable selection for clustering, and also demonstrate the benefits of Bayesian model averaging (BMA) in place of BMS. Through an array of simulation examples and well-studied examples from cancer transcriptomics, we show that our method performs competitively with the current state-of-the-art, while also offering computational benefits. We apply our approach to reverse-phase protein array (RPPA) data from The Cancer Genome Atlas (TCGA) in order to perform a pan-cancer proteomic characterisation of 5157 tumour samples. We have implemented our approach, together with the original SUGS algorithm, in an open-source R package named sugsvarsel, which accelerates analysis by performing intensive computations in C++ and provides automated parallel processing. The R package is freely available from: https://github.com/ococrook/sugsvarsel.

Retroviruses integrate into a shared, non-palindromic DNA motif.

Many DNA-binding factors, such as transcription factors, form oligomeric complexes with structural symmetry that bind to palindromic DNA sequences1. Palindromic consensus nucleotide sequences are also found at the genomic integration sites of retroviruses2-6 and other transposable elements7-9, and it has been suggested that this palindromic consensus arises as a consequence of the structural symmetry in the integrase complex2,3. However, we show here that the palindromic consensus sequence is not present in individual integration sites of human T-cell lymphotropic virus type 1 (HTLV-1) and human immunodeficiency virus type 1 (HIV-1), but arises in the population average as a consequence of the existence of a non-palindromic nucleotide motif that occurs in approximately equal proportions on the plus strand and the minus strand of the host genome. We develop a generally applicable algorithm to sort the individual integration site sequences into plus-strand and minus-strand subpopulations, and use this to identify the integration site nucleotide motifs of five retroviruses of different genera: HTLV-1, HIV-1, murine leukaemia virus (MLV), avian sarcoma leucosis virus (ASLV) and prototype foamy virus (PFV). The results reveal a non-palindromic motif that is shared between these retroviruses.

Robustness of MEK-ERK Dynamics and Origins of Cell-to-Cell Variability in MAPK Signaling.

Cellular signaling processes can exhibit pronounced cell-to-cell variability in genetically identical cells. This affects how individual cells respond differentially to the same environmental stimulus. However, the origins of cell-to-cell variability in cellular signaling systems remain poorly understood. Here, we measure the dynamics of phosphorylated MEK and ERK across cell populations and quantify the levels of population heterogeneity over time using high-throughput image cytometry. We use a statistical modeling framework to show that extrinsic noise, particularly that from upstream MEK, is the dominant factor causing cell-to-cell variability in ERK phosphorylation, rather than stochasticity in the phosphorylation/dephosphorylation of ERK. We furthermore show that without extrinsic noise in the core module, variable (including noisy) signals would be faithfully reproduced downstream, but the within-module extrinsic variability distorts these signals and leads to a drastic reduction in the mutual information between incoming signal and ERK activity.

A graph theoretical approach to data fusion.

The rapid development of high throughput experimental techniques has resulted in a growing diversity of genomic datasets being produced and requiring analysis. Therefore, it is increasingly being recognized that we can gain deeper understanding about underlying biology by combining the insights obtained from multiple, diverse datasets. Thus we propose a novel scalable computational approach to unsupervised data fusion. Our technique exploits network representations of the data to identify similarities among the datasets. We may work within the Bayesian formalism, using Bayesian nonparametric approaches to model each dataset; or (for fast, approximate, and massive scale data fusion) can naturally switch to more heuristic modeling techniques. An advantage of the proposed approach is that each dataset can initially be modeled independently (in parallel), before applying a fast post-processing step to perform data integration. This allows us to incorporate new experimental data in an online fashion, without having to rerun all of the analysis. We first demonstrate the applicability of our tool on artificial data, and then on examples from the literature, which include yeast cell cycle, breast cancer and sporadic inclusion body myositis datasets.

Plasma proteome analysis in HTLV-1-associated myelopathy/tropical spastic paraparesis.

BACKGROUND: Human T lymphotropic virus Type 1 (HTLV-1) causes a chronic inflammatory disease of the central nervous system known as HTLV-1-associated myelopathy/tropical spastic paraparesis (HAM) which resembles chronic spinal forms of multiple sclerosis (MS). The pathogenesis of HAM remains uncertain. To aid in the differential diagnosis of HAM and to identify pathogenetic mechanisms, we analysed the plasma proteome in asymptomatic HTLV-1 carriers (ACs), patients with HAM, uninfected controls, and patients with MS. We used surface-enhanced laser desorption-ionization (SELDI) mass spectrometry to analyse the plasma proteome in 68 HTLV-1-infected individuals (in two non-overlapping sets, each comprising 17 patients with HAM and 17 ACs), 16 uninfected controls, and 11 patients with secondary progressive MS. Candidate biomarkers were identified by tandem Q-TOF mass spectrometry. RESULTS: The concentrations of three plasma proteins--high [ß2-microglobulin], high [Calgranulin B], and low [apolipoprotein A2]--were specifically associated with HAM, independently of proviral load. The plasma [ß2-microglobulin] was positively correlated with disease severity. CONCLUSIONS: The results indicate that monocytes are activated by contact with activated endothelium in HAM. Using ß2-microglobulin and Calgranulin B alone we derive a diagnostic algorithm that correctly classified the disease status (presence or absence of HAM) in 81% of HTLV-1-infected subjects in the cohort.