Search for: All records

Abstract To understand genome evolution in a group of microbes, we need to know the timing of events such as duplications, deletions and horizontal transfers. A common approach is to perform a gene-tree / species-tree reconciliation. While a number of software packages perform this type of analysis, none are geared toward a complete reconstruction for all families in an entire clade. Here we describe an update to the xenoGI software package which allows users to perform such an analysis using the newly developed DTLOR (duplication-transfer-loss-origin-rearrangement) reconciliation model starting from genome sequences as input. 

BackgroundAnalyses of microbial evolution often use reconciliation methods. However, the standard duplication-transfer-loss (DTL) model does not account for the fact that species trees are often not fully sampled and thus, from the perspective of reconciliation, a gene family may enter the species tree from the outside. Moreover, within the genome, genes are often rearranged, causing them to move to new syntenic regions. ResultsWe extend the DTL model to account for two events that commonly arise in the evolution of microbes:originof a gene from outside the sampled species tree andrearrangementof gene syntenic regions. We describe an efficient algorithm for maximum parsimony reconciliation in this new DTLOR model and then show how it can be extended to account for non-binary gene trees to handle uncertainty in gene tree topologies. Finally, we describe preliminary experimental results from the integration of our algorithm into the existing xenoGI tool for reconstructing the histories of genomic islands in closely related bacteria. ConclusionsReconciliation in the DTLOR model can offer new insights into the evolution of microbes that is not currently possible under the DTL model. 

Discrete optimization problems arise in many biological contexts and, in many cases, we seek to make inferences from the optimal solutions. However, the number of optimal solutions is frequently very large and making inferences from any single solution may result in conclusions that are not supported by other optimal solutions. We describe a general approach for efficiently (polynomial time) and exactly (without sampling) computing statistics on the space of optimal solutions. These statistics provide insights into the space of optimal solutions that can be used to support the use of a single optimum (e.g., when the optimal solutions are similar) or justify the need for selecting multiple optima (e.g., when the solution space is large and diverse) from which to make inferences. We demonstrate this approach on two well-known problems and identify the properties of these problems that make them amenable to this method. 

Predicting the secondary structure of RNA is an important problem in molecular biology, providing insights into the function of non-coding Rn As and with broad applications in understanding disease, the development of new drugs, among others. Combinatorial algorithms for predicting RNA foldings can generate an exponentially large number of equally optimal foldings with respect to a given optimization criterion, making it difficult to determine how well any single folding represents the entire space. We provide efficient new algorithms for providing insights into this large space of optimal RNA foldings and a research software tool, toRNAdo, that implements these algorithms. 

Phylogenetic reconciliation is a fundamental method in the study of pairs of coevolving species. This paper provides an overview of the underlying theory of reconciliation in the context of host-symbiont cophylogenetics, identifying some of the major challenges to users of these methods, such as selecting event costs and selecting representative reconciliations. Next, recent advances to address these challenges are discussed followed by a discussion of several established and recent software tools.