Sparse relative risk regression models.

Clinical studies where patients are routinely screened for many genomic features are becoming more routine. In principle, this holds the promise of being able to find genomic signatures for a particular disease. In particular, cancer survival is thought to be closely linked to the genomic constitution of the tumor. Discovering such signatures will be useful in the diagnosis of the patient, may be used for treatment decisions and, perhaps, even the development of new treatments. However, genomic data are typically noisy and high-dimensional, not rarely outstripping the number of patients included in the study. Regularized survival models have been proposed to deal with such scenarios. These methods typically induce sparsity by means of a coincidental match of the geometry of the convex likelihood and a (near) non-convex regularizer. The disadvantages of such methods are that they are typically non-invariant to scale changes of the covariates, they struggle with highly correlated covariates, and they have a practical problem of determining the amount of regularization. In this article, we propose an extension of the differential geometric least angle regression method for sparse inference in relative risk regression models. A software implementation of our method is available on github (https://github.com/LuigiAugugliaro/dgcox).

Antibiotic Treatment in the First Week of Life Impacts the Growth Trajectory in the First Year of Life in Term Infants.

OBJECTIVE: Antibiotic treatment in early life appears to increase the risk for childhood overweight and obesity. So far, the association between antibiotics administrated specifically during the first week of life and growth has not been studied. Therefore, we studied the association between growth and antibiotics, given in the first week of life and antibiotic courses later in the first year of life. METHOD: A prospective observational birth cohort of 436 term infants with 151 receiving broad-spectrum antibiotics for suspected neonatal infection (AB+), and 285 healthy controls (AB-) was followed during their first year. Weight, height, and additional antibiotic courses were collected monthly. A generalized-additive-mixed-effects model was used to fit the growth data. Growth curve estimation was controlled for differences in sex, gestational age, delivery mode, exclusive breast-feeding, tobacco exposure, presence of siblings, and additional antibiotic courses. RESULTS: Weight-for-age and length-for-age increase was lower in AB+ compared with AB- (P < 0.0001), resulting in a lower weight and length increase 6.26 kg (standard error [SE] 0.07 kg) and 25.4 cm (SE 0.27 cm) versus 6.47 kg (SE 0.06 kg) and 26.4 cm (SE 0.21 cm) (P < 0.05 and P < 0.005, respectively) in the first year of life. Approximately 30% of the children in both groups received additional antibiotic course(s) in their first year, whereafter additional weight gain of 76 g per course was observed (P = 0.0285). CONCLUSIONS: Decreased growth was observed after antibiotics in the first week of life, whereas increased growth was observed after later antibiotic course(s) in term born infants in the first year of life. Therefore, timing of antibiotics may determine the association with growth.

De novo construction of polyploid linkage maps using discrete graphical models.

MOTIVATION: Linkage maps are used to identify the location of genes responsible for traits and diseases. New sequencing techniques have created opportunities to substantially increase the density of genetic markers. Such revolutionary advances in technology have given rise to new challenges, such as creating high-density linkage maps. Current multiple testing approaches based on pairwise recombination fractions are underpowered in the high-dimensional setting and do not extend easily to polyploid species. To remedy these issues, we propose to construct linkage maps using graphical models either via a sparse Gaussian copula or a non-paranormal skeptic approach. RESULTS: We determine linkage groups, typically chromosomes, and the order of markers in each linkage group by inferring the conditional independence relationships among large numbers of markers in the genome. Through simulations, we illustrate the utility of our map construction method and compare its performance with other available methods, both when the data are clean and contain no missing observations and when data contain genotyping errors. Our comprehensive map construction method makes full use of the dosage SNP data to reconstruct linkage map for any bi-parental diploid and polyploid species. We apply the proposed method to three genotype datasets: barley, peanut and potato from diploid and polyploid populations. AVAILABILITY AND IMPLEMENTATION: The method is implemented in the R package netgwas which is freely available at https://cran.r-project.org/web/packages/netgwas. SUPPLEMENTARY INFORMATION: Supplementary data are available at Bioinformatics online.

Stability estimation of autoregulated genes under Michaelis-Menten-type kinetics.

Feedback loops are typical motifs appearing in gene regulatory networks. In some well-studied model organisms, including Escherichia coli, autoregulated genes, i.e., genes that activate or repress themselves through their protein products, are the only feedback interactions. For these types of interactions, the Michaelis-Menten (MM) formulation is a suitable and widely used approach, which always leads to stable steady-state solutions representative of homeostatic regulation. However, in many other biological phenomena, such as cell differentiation, cancer progression, and catastrophes in ecosystems, one might expect to observe bistable switchlike dynamics in the case of strong positive autoregulation. To capture this complex behavior we use the generalized family of MM kinetic models. We give a full analysis regarding the stability of autoregulated genes. We show that the autoregulation mechanism has the capability to exhibit diverse cellular dynamics including hysteresis, a typical characteristic of bistable systems, as well as irreversible transitions between bistable states. We also introduce a statistical framework to estimate the kinetics parameters and probability of different stability regimes given observational data. Empirical data for the autoregulated gene SCO3217 in the SOS system in Streptomyces coelicolor are analyzed. The coupling of a statistical framework and the mathematical model can give further insight into understanding the evolutionary mechanisms toward different cell fates in various systems.

Autonomous Metabolic Oscillations Robustly Gate the Early and Late Cell Cycle.

Eukaryotic cell division is known to be controlled by the cyclin/cyclin dependent kinase (CDK) machinery. However, eukaryotes have evolved prior to CDKs, and cells can divide in the absence of major cyclin/CDK components. We hypothesized that an autonomous metabolic oscillator provides dynamic triggers for cell-cycle initiation and progression. Using microfluidics, cell-cycle reporters, and single-cell metabolite measurements, we found that metabolism of budding yeast is a CDK-independent oscillator that oscillates across different growth conditions, both in synchrony with and also in the absence of the cell cycle. Using environmental perturbations and dynamic single-protein depletion experiments, we found that the metabolic oscillator and the cell cycle form a system of coupled oscillators, with the metabolic oscillator separately gating and maintaining synchrony with the early and late cell cycle. Establishing metabolism as a dynamic component within the cell-cycle network opens new avenues for cell-cycle research and therapeutic interventions for proliferative disorders.

Confidence intervals for intraclass correlation coefficients in variance components models.

Confidence intervals for intraclass correlation coefficients in agreement studies with continuous outcomes are model-specific and no generic approach exists. This paper provides two generic approaches for intraclass correlation coefficients of the form [Formula: see text] The first approach uses Satterthwaite's approximation and an F-distribution. The second approach uses the first and second moments of the intraclass correlation coefficient estimate in combination with a Beta distribution. Both approaches are based on the restricted maximum likelihood estimates for the variance components involved. Simulation studies are conducted to examine the coverage probabilities of the confidence intervals for agreement studies with a mix of small sample sizes. Two different three-way variance components models and balanced and unbalanced one-way random effects models are investigated. The proposed approaches are compared with other approaches developed for these specific models. The approach based on the F-distribution provides acceptable coverage probabilities, but the approach based on the Beta distribution results in accurate coverages for most settings in both balanced and unbalanced designs. A real agreement study is provided to illustrate the approaches.

Borrowing strength: a likelihood ratio test for related sparse signals.

MOTIVATION: Cancer biology is a field where the complexity of the phenomena battles against the availability of data. Often only a few observations per signal source, i.e. genes, are available. Such scenarios are becoming increasingly more relevant as modern sensing technologies generally have no trouble in measuring lots of channels, but where the number of subjects, such as patients or samples, is limited. In statistics, this problem falls under the heading 'large p, small n'. Moreover, in such situations the use of asymptotic analytical results should generally be mistrusted. RESULTS: We consider two cancer datasets, with the aim to mine the activity of functional groups of genes. We propose a hierarchical model with two layers in which the individual signals share a common variance component. A likelihood ratio test is defined for the difference between two collections of corresponding signals. The small number of observations requires a careful consideration of the bias of the statistic, which is corrected through an explicit Bartlett correction. The test is validated on Monte Carlo simulations, which show improved detection of differences compared with other methods. In a leukaemia study and a cancerous fibroblast cell line, we find that the method also works better in practice, i.e. it gives a richer picture of the underlying biology. AVAILABILITY: The MATLAB code is available from the authors or on http://www.math.rug.nl/stat/Software. CONTACT: e.c.wit@rug.nl d.bakewell@liv.ac.uk.

Markers of adenocarcinoma characteristic of the site of origin: development of a diagnostic algorithm.

PURPOSE: Patients with metastatic adenocarcinoma of unknown origin are a common clinical problem. Knowledge of the primary site is important for their management, but histologically, such tumors appear similar. Better diagnostic markers are needed to enable the assignment of metastases to likely sites of origin on pathologic samples. EXPERIMENTAL DESIGN: Expression profiling of 27 candidate markers was done using tissue microarrays and immunohistochemistry. In the first (training) round, we studied 352 primary adenocarcinomas, from seven main sites (breast, colon, lung, ovary, pancreas, prostate and stomach) and their differential diagnoses. Data were analyzed in Microsoft Access and the Rosetta system, and used to develop a classification scheme. In the second (validation) round, we studied 100 primary adenocarcinomas and 30 paired metastases. RESULTS: In the first round, we generated expression profiles for all 27 candidate markers in each of the seven main primary sites. Data analysis led to a simplified diagnostic panel and decision tree containing 10 markers only: CA125, CDX2, cytokeratins 7 and 20, estrogen receptor, gross cystic disease fluid protein 15, lysozyme, mesothelin, prostate-specific antigen, and thyroid transcription factor 1. Applying the panel and tree to the original data provided correct classification in 88%. The 10 markers and diagnostic algorithm were then tested in a second, independent, set of primary and metastatic tumors and again 88% were correctly classified. CONCLUSIONS: This classification scheme should enable better prediction on biopsy material of the primary site in patients with metastatic adenocarcinoma of unknown origin, leading to improved management and therapy.

Identification from public data of molecular markers of adenocarcinoma characteristic of the site of origin.

Patients presenting with metastatic adenocarcinoma of unknown origin are a common clinical problem. Their optimal management and therapy are facilitated by identification of the primary site, yet histologically these tumors are almost identical. Better tumor markers are needed to enable the assignment of metastases to likely sites of origin. In this study, hierarchical clustering of public serial analysis of gene expression data showed that adenocarcinomas and their metastases cluster according to their site of origin. A novel bioinformatic approach was developed to exploit the differences between these clusters, using diverse sources: public expression data from serial analysis of gene expression and digital differential display; and the published literature, including microarray studies. Sixty-one candidate tumor markers with expression predicted to be characteristic of the site of origin were identified. Eleven genes were tested by reverse transcription-PCR in primary adenocarcinomas from a range of sites, and seven (64%) were site-restricted. Analysis of public gene expression data sets is a powerful method for the identification of clinically relevant tumor markers.