Short Abstract: Decades ago along with the usage of compositional methods, the idea of amelioration whereby the base DNA composition of the transferred genes from a donor undergoes nucleotide substitutions over time and reflects similarly in DNA composition to the recipient genome, was introduced. Since its introduction, few studies have been done and little is known about its process. With the usage of oligonucleotide frequency and Verhulst Modeling, we hereby propose a method to simulate the amelioration process as well as model it mathematically. The mathematical model will be used to answer questions related to amelioration such as time of insert of mobile genetic elements as well as factors influencing amelioration. As foreign inserts, several known genomic islands (GI) were used to model their amelioration process towards compositional profiles of genomes of organisms representing distant taxa and different GC content, i.e. Bacillus subtilis 168, Pseudomonas aeruginosa PA01, Escherichia coli K12, Xylella fastidiosa 9a5c and Streptomyces griseus NBRC 13350.Compositional methods were used on each combination of GI (tester) and recipient (target) and an amelioration model was then derived and fitted to the standard Verhulst model, which in general fits well to majority of the combination of tester and target. The parameter within the model is well suited for the simulation data and represents biologically meaningful charateristics such as rate and selectivity of mutations and graduate merging of the insert's OU profiles with those of the host genomes that would stabilize at some level of pattern similarity which depends on the composition of tester and target sequence.

Short Abstract: The evolution and transmission of antimicrobial resistant bacteria is a major threat to public health. Escherichia coli sequence type ST131 (O25:H4) is a pandemic multidrug-resistant pathogen and is associated with the CTX-M-15 extended-spectrum beta-lactamase. PFGE- and MLST-based approaches have much less power compared to genome-sequencing to resolve pathogen evolution due to the lower number of biomarkers. Moreover, standard methods for partitioning variation into discrete categories can miss recombination events. In this study spanning 2005-2011, 96 E. coli isolates were collected from hospitals (n=8), the community (n=2) and from residents in eleven long-term care facilities across Ireland (n=11) and one single care facility (n=69 from 63 residents). 90 of the 96 (94%) were ST131: these were compared with 11 other Indian/UK ST131 for evolutionary context (n=101). After Illumina Hiseq sequencing, reading-mapping and conservative variant calling, a total 8,687 SNPs were discovered in these 101 genomes. Phylogenomic and model-based analyses indicated five discrete populations, but only when non-recombining SNPs were excluded. Subsequent evaluations of haplotype flow within and between populations over time indicated periodical displacement of distinct populations within little recombination overall between groups, including one displacement event within a single resident. Haplotype-based inference of recombination indicated that a binary differentiation of mutations in ones arising by recombination or not across the whole collection limited inferences because most recombination events were between isolate pairs. This scheme identified the transferred genes as transposases, bacteriophage elements and pathogenicity islands. This highlighted the importance of sensitive approaches for untangling historical patterns of horizontal gene exchange.

Short Abstract: Tularemia outbreaks, caused by a subpopulation of Francisella tularensis has been reported to be emerging in Continental Western Europe with Spain reporting its first case in 1996. Little is known about the underlying population genetics of the bacterium. High resolution clone tracking would improve our understanding of the genetic diversity in the region and offer clues into the evolution of the bacteria.
Sixty eight whole genome sequences of F. tularensis from Western Europe, isolated between 1947 and 2013 were analyzed and compared with sequences of global origin. Based on SNP markers obtained from the genomic analysis, another 155 isolates were genotyped and several genetic parameters were tested.
All isolates from Western Europe belonged to a specific clade of Francisella which has only been isolated in the study region. Two genetic branches were identified: one exhibited a star phylogeny and was observed throughout the region and another was geographically restricted to the Alps region. Mutation rate of 0.4 mutations per year was calculated. Comparison with historical outbreaks revealed fixed endemic regions.
Presence of single genetically homogenous clade specific to the study region points to a recent clonal spread of the pathogen in Western Europe. The star phylogeny is consistent with rapid clonal expansion of the clade. Limited genetic diversity over 67 year time scale could be attributed to a very low natural mutation rate. The difference between clades in the geographic spread, number of cases and in the dN/dS ratio points towards distinct fitness levels and possibly different ecologies.

Short Abstract: Although individuals in a genome wide association study might not be related to each other, there can exist a distant relationship between these individuals. This cryptic relationship violates the assumption regarding the independence of the subject genomes, causing results to be both false positive and false negative. Hence we present a method to correct these cryptic relationships.
We start by accurately detecting distant relationship using and expectation maximization (EM) algorithm for identity coefficients (a refinement of kinship coefficient). After this, we compute the kinship coefficient and apply a kinship-corrected association test.
We analyze and show that genome simulated from Wright-Fisher pedigree, our approach converges quickly and is accurate requiring relatively small number of sites. To assess the kinship-corrected association test, we simulated individuals from deep pedigrees and drew one site which recessively determined the disease status. Once we estimated the kinship coefficient through our method, we kinship-adjusted test, where the results were favorable to our method compared to the state-of-the-art covariance-based approach.
Use of our method to find cryptic relationships and for corrected association tests is advantageous, because it is easy to interpret through the use of identity states as latent variables, and its results provides state-of-the-accuracy when compared to models that only consider kinship coefficients.

Short Abstract: Lentil (Lens culinaris) is becoming an increasingly important food crop, but due to its relatively large genome (~4Gb) there have been few genomic resources available to the breeding community until recently. Wild species within the genus Lensare being used within existing breeding programs as sources of new genetic variation for important traits such as disease resistance. To better understand the relationships and degree of genetic variability among species of interest to the breeding program, as well as create markers for identifying introgressions from wild species, we performed genotyping by sequencing (GBS) on 60 lines from seven species: L. culinaris, L. orientalis, L. tomentosus, L. lamottei, L. odemensis, L. ervoides and L. nigricans. We made use of standard tools within the bioinformatics community to put together our own pipeline for handling GBS data and subsequent variant calling, generating 5389 high-quality SNPs. Phylogenetic analysis resulted in the characterization of four gene pools and panels of markers associated with each species, as well as identification of germplasm which had been previously misclassified. Biases due to reference selection and DNA quality were observed, as well as run-to-run variation.

Short Abstract: The Great Oxidation Event (GOE) w2.4 billion years ago resulted from the accumulation of oxygen by the ancestors of cyanobacteria. Cyanobacteria continue to play a significant role in primary production and in regulating the global marine and limnic nitrogen cycles. Relatively little is known, however, about the evolutionary history and gene content of primordial cyanobacteria. To address these issues, we used protein similarity networks, containing proteomes from 48 cyanobacteria as the test group, and reference proteomes from 84 microbes representing four distinct metabolic groups from most reducing to most oxidizing: methanogens, obligate anaerobes (nonmethanogenic), facultative aerobes, and obligate aerobes. These four metabolic groups represent extant bioinformatic proxies for ancient redox chemistries, extending from an anoxic origin through the GOE and ultimately to obligate aerobes. Analysis of the network metric degree showed a strong relationship between cyanobacteria and obligate anaerobes, from which cyanobacteria presumably arose, for core functions that include translation, photosynthesis, energy conservation, and environmental interactions. These data were used to reconstruct primordial functions in cyanobacteria that included nine gene families involved in photosynthesis,
hydrogenases, and proteins involved in defense from environmental stress. The presence of 60% of these
genes in both reaction center I (RC-I) and RC-II-type bacteria may be explained by selective loss of either RC in the evolutionary history of some photosynthetic lineages. Finally, the network reveals that cyanobacteria occupy a unique position among prokaryotes as a hub between anaerobes and obligate aerobes.

Short Abstract: The broad and inexpensive availability of modern next-generation sequencing and genotyping technologies have led to a wealth of data and analytical methods for population genetics research. In the popular statistical and mathematical computing platform R alone, there are now dozens of packages available for analyzing and visualizing population genetics data. However, this organically grown wealth of methods and packages, combined with the exponential growth of datasets, has also created challenges for researchers to take full advantage of these resources. It can be difficult to know which R packages are best used, and many packages do not interoperate well. A common base class that provides efficient storage of genetic data and promotes interoperability remains lacking even though the need was identified years ago. Current implementations often do not scale well to the kind of large volume datasets that are increasingly common. Creating complex analysis workflows that need to pass data, metadata, and other state information from one package to another can be challenging. To address these gaps, we held the Population Genetics in R Hackathon at, and sponsored by the National Evolutionary Synthesis Center (NESCent), located in Durham, NC. We designed the event to target interoperability, scalability, and workflow building challenges among the many population genetics R packages that already exist, and to ultimately help foster an interoperating ecosystem of tools and resources for both users and researcher-developers. Here we highlight key outcomes of the event, which are all open-source and freely available. For details, see https://github.com/NESCent/r-popgen-hackathon/.

Short Abstract: Background: Chromosomal copy number variants (CNVs) associated with schizophrenia have been shown to have the broad risk for pediatric populations with neurodevelopmental disorders more generally. Method: A selected subgroup of CNVs found elevated in a clinical childhood onset schizophrenia population were screen ed within a sample of 40,000 pediatric patents whose pre scored records & DNA were available in the CHOP cohort study. Neurodevelopmental and other disorders were compared with 5:1 matched control for each CNV who were not carrying any of these CNVs, to see prevalence of any disorder was sufficiently high to inform genetic counseling. Results: Six of these CNVs were detected at rates estimated from other large control population studies, Having any of the 6 CNVs measured collectively showed an increase four of the 14 disease categories examined: congenital deficit, surgery, mental disorder and digestive system disorder), and a mental disorder (primarily developmental delay). The most prominent associations were between 22q11duplication, and gastroesophageal reflux disorder which was more likely to be present in the presence of developmental delay . The other CNV surviving correct on was 16p11.2del for which both mental and nervous system 16p11 deletion were significant . Conclusion: A broader concept of overall clinical penetrance may be of importance for genetic counseling. This is the first report of CNV association of clinical significance.

Short Abstract: Beta-splitting trees, as described by Aldous, is a distribution on binary tree structures defined by a discrete time continuous state process. The process is described by a hierarchy of "splitting points" which partitions the unit interval. A set of individuals live on the interval, and each splitting event thus partitions the individuals. For a finite set of individuals, and for particular choices of splitting distributions, integrating out the splitting points recovers many familiar distributions on trees, for example the uniform distribution on trees, and also the distribution on tree structures induced by Kingman's Coalescent. We define consistent time variables using beta-splitting for use in hierarchical clustering. This choice of time variable induces a relationship between cluster size and cluster instantiation time, giving a model in which older clusters/populations are likely to be larger. We perform inference using this model on a tumor subclonal reconstruction task, where this dependence between time and population size is particularly relevant.