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.     

Growth Phase-Dependent Regulation of Vsr Endonuclease May Contribute to 5-Methylcytosine Mutational Hot Spots in Escherichia coli

Using rabbit polyclonal antibodies, we have shown that the Dcm cytosine methylase of Escherichia coli is maintained at a constant level during cell growth, while Vsr endonuclease levels are growth phase dependent. Decreased production of Vsr relative to Dcm during the log phase may contribute substantially to the mutability of 5-methylcytosine.     

Effects of Amino Acid Substitutions at the Active Site in Escherichia Coli β-Galactosidase

Forty-nine amino acid substitutions were made at four positions in the Escherichia coli enzyme β-galactosidase; three of the four targeted amino acids are thought to be part of the active site. Many of the substitutions were made by converting the appropriate codon in lacZ to an amber codon, and using one of 12 suppressor strains to introduce the replacement amino acid. Glu-461 and Tyr-503 were replaced, independently, with 13 amino acids. All 26 of the strains containing mutant enzymes are Lac(-). Enzyme activity is reduced to less than 10% of wild type by substitutions at Glu-461 and to less than 1% of wild type by substitutions at Tyr-503. Many of the mutant enzymes have less than 0.1% wild-type activity. His-464 and Met-3 were replaced with 11 and 12 amino acids, respectively. Strains containing any one of these mutant proteins are Lac(+). The results support previous evidence that Glu-461 and Tyr-503 are essential for catalysis, and suggest that His-464 is not part of the active site. Site-directed mutagenesis was facilitated by construction of an f1 bacteriophage containing the complete lacZ gene on a single EcoRI fragment.     

Maternal factors in onset of Huntington disease.

Analyses of father-offspring and mother-offspring similarity in onset age suggest that nuclear genes account for a significant portion of the modification of onset age in Huntington disease. The effects of non-nuclear modifiers are supported by the finding that the offspring of affected women have significantly older mean ages of onset than offspring of affected men irrespective of the onset age in the parent. The absence of increased father-daughter similarity indicates that modification is not X-linked. The absence of reproductive advantage for late-onset individuals and the absence of a multigenerational maternal-lineage effect suggest that the modifying effect of the sex of the affected parent occurs in a single parental generation. Offspring of affected women with onset between ages 35 and 49 had a significantly older mean onset age than their mothers. This suggests that a protective effect may be conferred upon the offspring of affected women.     

Inverse relationship between age at onset of Huntington disease and paternal age suggests involvement of genetic imprinting.

It is well recognized that age at onset of Huntington disease (HD) is strongly influenced by the sex of the affected parent, and this has lead to suggestions that genetic imprinting or maternal specific factors may play a role in the expression of the disease. This study evaluated maternal and paternal ages, birth order, parental age at onset, and sex of the affected parent and grandparent in 1,764 patients in the National HD Roster by using linear-regression techniques which incorporated a weighted least-squares approach to accommodate the correlation among siblings. It was found that paternal age is negatively associated with age at onset of HD, particularly among subjects who inherit the mutant gene from grandfathers. Apparent associations between age at onset and birth order and between age at onset and maternal age were not significant after adjustment for paternal age. The paternal age effect is strongest among juvenile-onset cases and individuals with anticipation of greater than or equal to 10 years, although it is detectable across the entire age-at-onset distribution. The tendency for older fathers, including those not transmitting the HD gene, to have affected offspring with early-onset disease may be consistent with a gene imprinting mechanism involving DNA methylation. Because paternal age in unaffected fathers is also a significant determinant of age at onset, methylation in this context might involve HD modifier genes or the normal HD allele.     

A Familial Factor Independent of CAG Repeat Length Influences Age at Onset of Machado-Joseph Disease

Machado-Joseph disease (MJD) is a late-onset, progressive, neurodegenerative disorder caused by the expansion of an unstable trinucleotide (CAG) repeat sequence in a novel gene (MJD1) on chromosome 14. Previous studies showed that age at onset is negatively correlated with the number of CAG repeat units, but only part of the variation in onset age is explained by CAG repeat length. Ages at onset and CAG repeat lengths of 136 MJD patients from 23 kindreds of Portuguese descent were analyzed, to determine whether familial factors independent of CAG repeat length modulate age at onset of MJD. Correlation among sibs for onset age adjusted for CAG repeat length was .43, which indicates that an environmental or genetic factor common to sibs influences onset age. Positive correlations were also observed for avuncular (r = .22) and first-cousin pairs (r = .28), which supports the hypothesis that a genetic factor is influencing age at onset. Commingling analysis of onset ages adjusted for CAG repeat length identified three distributions in this population of affected individuals. Further studies of a much larger sample are needed to determine whether these distributions represent the influence of a genetic or environmental factor.     

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.     

Bacterial colonization of Hydra hatchlings follows a robust temporal pattern

Animals are colonized by complex bacterial communities. The processes controlling community membership and influencing the establishment of the microbial ecosystem during development are poorly understood. Here we aimed to explore the assembly of bacterial communities in Hydra with the broader goal of elucidating the general rules that determine the temporal progression of bacterial colonization of animal epithelia. We profiled the microbial communities in polyps at various time points after hatching in four replicates. The composition and temporal patterns of the bacterial communities were strikingly similar in all replicates. Distinct features included high diversity of community profiles in the first week, a remarkable but transient adult-like profile 2 weeks after hatching, followed by progressive emergence of a stable adult-like pattern characterized by low species diversity and the preponderance of the Betaproteobacterium Curvibacter. Intriguingly, this process displayed important parallels to the assembly of human fecal communities after birth. In addition, a mathematical modeling approach was used to uncover the organizational principles of this colonization process, suggesting that both, local environmental or host-derived factor(s) modulating the colonization rate, as well as frequency-dependent interactions of individual bacterial community members are important aspects in the emergence of a stable bacterial community at the end of development.