Selecting high-dimensional mixed graphical models using minimal AIC or BIC forests

Background  

Chow and Liu showed that the maximum likelihood tree for multivariate discrete distributions may be found using a maximum weight spanning tree algorithm, for example Kruskal's algorithm. The efficiency of the algorithm makes it tractable for high-dimensional problems.     

Respiratory response to chemical stimuli and exercise capacity under conditions of acute hypoxia in elite mountain climbers]

To investigate the factors that modulate exercise performance at extreme altitude, the role of the following variables was analyzed in 16 climbers: 1) ventilatory response to chemical stimuli (hypoxia and hypercapnia); and, 2) maximum exercise performance while breathing room air and during acute hypoxia (F1O2, 0.11). Seven climbers (elite climbers, AE) had previously ascended to 8,000 m or more above sea level, and 9 (A) had never achieved such extreme altitude. Then healthy sedentary subjects (C) of similar age (31.1 +/- 6.0 SD years) were used as control group. Elite climbers showed higher ventilatory responses to both transient hypoxia (-0.49 +/- 0.13 L x min-1 x %-1) (p less than 0.05) and progressive hypoxia (-0.47 +/- 0.13 L x min-1 x %-1) than C (-0.33 +/- 0.14 and -0.30 +/- 0.15 L x min-1 x %-1, respectively). By contrast, no differences were observed between the two groups of climbers. The ventilatory response to hypercapnia was higher in AE (3.04 +/- 1.03 L x min-1 mmHg-1) compared to A (1.85 +/- 0.73 L x min-1 mmHg-1) (p less than 0.05) but similar to that observed in C. Breathing 11% O2, maximum workload and oxyhemoglobin desaturation during maximum exercise were similar in both groups of climbers. Additionally, the ventilatory response to hypoxia did not correlate with maximum workload (F1O2, 0.11), maximal ventilation during exercise (F1O2, 0.11), nor with the altitude score. The present study supports previous reports that inform about the role of the ventilatory response to hypoxia in the exercise performance at high altitude.(ABSTRACT TRUNCATED AT 250 WORDS)     

Order-disorder phenomena in myelinated nerve sheaths. II. The structure of myelin in native and swollen rat sciatic nerves and in the course of myelinogenesis

The algorithm described in the accompanying paper was applied to X-ray scattering experiments performed with rat sciatic nerves, either as a function of the age of the animal (4 to 30 days), or with adult nerves swollen in non-isotonic media. The results were all consistent with the model of disorder used in the theoretical treatment. The algorithm leads, in one step, from the data to the numerical values of the parameters, avoiding all intermediate manipulation. For each experiment a variety of parameters was determined: the average D and the variance sigma 2D of the repeat distance, the average number [N] of motifs per crystallite, the set [idiff(h/D)], which defines the diffuse scattering, the fraction alphaloose of myelin that does not belong to the compact sheaths, and the set [imotif (k/2D)], which suffices to define the continuous intensity curve of the motif imotif(s). Note the remarkable wealth of information, especially by contrast with conventional analyses which, as a rule, only yield the values of D and of the set [imotif(h/D)] (insufficient to determine the function imotif(s]. The function imotif(s) and the parameters D and sigma D (and thus the local structure of the myelin sheaths) were shown to be almost invariant in the course of myelinogenesis; what varies is mainly the total amount of myelin in the nerve and the number of membranes per sheath. Swelling agents have a dramatic influence on the X-ray scattering spectra, but in spite of the conspicuous variation of D, sigma D and [N] the structure of the motif is invariant. The structure of the motif was shown to be quite different in the native and in the swollen samples; the stacking disorder appears to involve mainly the cytoplasmic space in native myelin, the external space in swollen nerves. The very notion of electron density profile, when disorder is present, is discussed. Two criteria were proposed to select the "best" signs of the reflections: two sets came out at almost the same rank, one corresponding to Caspar & Kirschner's the other to Worthington & McIntosh's proposals, neither of which can be ruled out according to the criteria used in this work.     

A mechanistic spatio-temporal modeling of COVID-19 data

Understanding the evolution of an epidemic is essential to implement timely and efficient preventive measures. The availability of epidemiological data at a fine spatio-temporal scale is both novel and highly useful in this regard. Indeed, having geocoded data at the case level opens the door to analyze the spread of the disease on an individual basis, allowing the detection of specific outbreaks or, in general, of some interactions between cases that are not observable if aggregated data are used. Point processes are the natural tool to perform such analyses. We analyze a spatio-temporal point pattern of Coronavirus disease 2019 (COVID-19) cases detected in Valencia (Spain) during the first 11 months (February 2020 to January 2021) of the pandemic. In particular, we propose a mechanistic spatio-temporal model for the first-order intensity function of the point process. This model includes separate estimates of the overall temporal and spatial intensities of the model and a spatio-temporal interaction term. For the latter, while similar studies have considered different forms of this term solely based on the physical distances between the events, we have also incorporated mobility data to better capture the characteristics of human populations. The results suggest that there has only been a mild level of spatio-temporal interaction between cases in the study area, which to a large extent corresponds to people living in the same residential location. Extending our proposed model to larger areas could help us gain knowledge on the propagation of COVID-19 across cities with high mobility levels.     

Structure of human serum lipoproteins in solution. II. Small-angle x-ray scattering study of HDL and LDL.

Two human serum lipoprotein particles, HDL₃ and LDL, were studied in solution in solvents of variable density (NaBr in water) by small-angle X-ray scattering using a position-sensitive proportional counter. The data were analysed using the theoretical approach outlined in the accompanying paper (Luzzati et al., 1976). The structures of the particles were found to be independent of the salt concentration of the solution (i.e. the particles are impenetrable to NaBr). Density heterogeneities are negligible and size and shape heterogeneities appear to be small.The particle structures could be quantitatively described in terms of a set of parameters and of a few one-dimensional functions. The parameters are the volume, radius of gyration and surface area of the shape functions; the second moment and square average of the electron density contrast at buoyancy; the electron density level, volume, radius of gyration and surface area of the hydrocarbon and polar regions. The one-dimensional functions are: the distribution of chords, the spherical average of the shape function and of the electron density at buoyancy, and the fraction of each spherical shell occupied by the hydrocarbon and polar regions. These parameters and functions are internally consistent and agree with the chemical data confirming the assumptions made in their derivation.The results are compatible with the shape of the particle being compact and quasi-spherical although with deeply convoluted surfaces. They also indicate that the outer layers of the particles are occupied by the proteins and the polar groups of phospholipids and free cholesterol, and the cores by neutral lipids. The maximum diameters of the particles are 130Å and 280Å for HDL₃ and LDL, respectively, while the hydrocarbon cores have diameters of 80Å and 230Å, respectively. The solvent is considered to penetrate to 25Å from the center of the HDL₃ particle with a minimum solvation at a radius of 45Å. In the case of LDL, the solvent penetrates to 55Å from the center of the particle. The lipids in the cores of the particles, particularly the cholesterol esters, appear to display a micelle-like organization with the steroid nuclei segregated in regions distinct from those occupied by the hydrocarbon chains.Although the data are consistent with several aspects of previously proposed models, they indicate that the structures of the HDL₃ and LDL particles are more complex than previously believed.     

Evolution of the capsid protein genes of foot-and-mouth disease virus: antigenic variation without accumulation of amino acid substitutions over six decades.

The genetic diversification of foot-and-mouth disease virus (FMDV) of serotype C over a 6-decade period was studied by comparing nucleotide sequences of the capsid protein-coding regions of viruses isolated in Europe, South America, and The Philippines. Phylogenetic trees were derived for VP1 and P1 (VP1, VP2, VP3, and VP4) RNAs by using the least-squares method. Confidence intervals of the derived phylogeny (significance levels of nodes and standard deviations of branch lengths) were placed by application of the bootstrap resampling method. These procedures defined six highly significant major evolutionary lineages and a complex network of sublines for the isolates from South America. In contrast, European isolates are considerably more homogeneous, probably because of the vaccine origin of several of them. The phylogenetic analysis suggests that FMDV CGC Ger/26 (one of the earliest FMDV isolates available) belonged to an evolutionary line which is now apparently extinct. Attempts to date the origin (ancestor) of the FMDVs analyzed met with considerable uncertainty, mainly owing to the stasis noted in European viruses. Remarkably, the evolution of the capsid genes of FMDV was essentially associated with linear accumulation of silent mutations but continuous accumulation of amino acid substitutions was not observed. Thus, the antigenic variation attained by FMDV type C over 6 decades was due to fluctuations among limited combinations of amino acid residues without net accumulation of amino acid replacements over time.     

The alternative oxidase pathway is involved in optimizing photosynthesis in Medicago truncatula infected by Fusarium oxysporum and Rhizoctonia solani

Phytopathogen infection alters primary metabolism status and plant development. The alternative oxidase (AOX) has been hypothesized to increase under pathogen attack preventing reductions, thus optimizing photosynthesis and growth. In this study, two genotypes of Medicago truncatula, one relatively resistant (Jemalong A17) and one susceptible (TN1.11), were infected with Fusarium oxysporum and Rhizoctonia solani. The in vivo foliar respiratory activities of the cytochrome oxidase pathway (COP) and the alternative oxidase pathway (AOP) were measured using the oxygen isotope fractionation. Gas exchange and photosynthesis-related parameters were measured and calculated together with antioxidant enzymes activities and organic acids contents. Our results show that the in vivo activity of AOX (v_alt) plays a role under fungal infection. When infected with R. solani, the increase of v_alt in A17 was concomitant to an increase in net assimilation, in mesophyll conductance, to an improvement in the maximum velocity of Rubisco carboxylation and to unchanged malate content. However, under F. oxysporum infection, the induced v_alt was accompanied by an enhancement in the antioxidant enzymes, superoxide dismutase (SOD; EC1.15.1.1), catalase (CAT; EC1.11.1.6) and guaiacol peroxidase (GPX; EC1.11.1.7), activities and to an unchanged tricarboxylic acid cycle intermediates. These results provide new insight into the role of the in vivo activity of AOX in coordinating primary metabolism interactions that, partly, modulate the relative resistance of M. truncatula to diseases caused by soil-borne pathogenic fungi.     

Geographical patterns of genetic variation in rosemary (Rosmarinus officinalis) in the Mediterranean basin

Climate changes during the Quaternary had important effects on the evolution of European plant species. The distribution of genetic variability in rosemary, a strictly Mediterranean species of reputed Plio‐Quaternary origin for which the diversification centre is hypothesized to be located in the western part of the Mediterranean basin, was investigated across the species range by using plastid microsatellites [plastid simple sequence repeat (cpSSR)] markers. Seven out of the 17 primer pairs screened were polymorphic, with up to four alleles, yielding a total of 17 size variants combined into ten haplotypes. A permutation test to investigate for geographical structure showed no significant differences between R_ST and G_ST, indicating that the species lacks geographical structure. Low correlation between genetic and geographical distances was shown by the Mantel test. Bayesian analysis identified two coancestry groups of populations. The distribution of genetic diversity supports the hypothesized origin in the western Mediterranean basin, and with the demographic expansion test indicates three different routes of migration: a northern route expanding along the northern side of the Mediterranean and two southern routes, one from west to east through North Africa and reaching Cyrenaica, and a second to the south‐west of the Iberian Peninsula, from where it came back to the south–central areas. © 2013 The Linnean Society of London, Botanical Journal of the Linnean Society, 2013 , 171 , 700–712.     

Enlarging a training set for genomic selection by imputation of un-genotyped animals in populations of varying genetic architecture

Background

The most common application of imputation is to infer genotypes of a high-density panel of markers on animals that are genotyped for a low-density panel. However, the increase in accuracy of genomic predictions resulting from an increase in the number of markers tends to reach a plateau beyond a certain density. Another application of imputation is to increase the size of the training set with un-genotyped animals. This strategy can be particularly successful when a set of closely related individuals are genotyped.

Methods

Imputation on completely un-genotyped dams was performed using known genotypes from the sire of each dam, one offspring and the offspring’s sire. Two methods were applied based on either allele or haplotype frequencies to infer genotypes at ambiguous loci. Results of these methods and of two available software packages were compared. Quality of imputation under different population structures was assessed. The impact of using imputed dams to enlarge training sets on the accuracy of genomic predictions was evaluated for different populations, heritabilities and sizes of training sets.

Results

Imputation accuracy ranged from 0.52 to 0.93 depending on the population structure and the method used. The method that used allele frequencies performed better than the method based on haplotype frequencies. Accuracy of imputation was higher for populations with higher levels of linkage disequilibrium and with larger proportions of markers with more extreme allele frequencies. Inclusion of imputed dams in the training set increased the accuracy of genomic predictions. Gains in accuracy ranged from close to zero to 37.14%, depending on the simulated scenario. Generally, the larger the accuracy already obtained with the genotyped training set, the lower the increase in accuracy achieved by adding imputed dams.

Conclusions

Whenever a reference population resembling the family configuration considered here is available, imputation can be used to achieve an extra increase in accuracy of genomic predictions by enlarging the training set with completely un-genotyped dams. This strategy was shown to be particularly useful for populations with lower levels of linkage disequilibrium, for genomic selection on traits with low heritability, and for species or breeds for which the size of the reference population is limited.