Biology subjects

Craft, D.

Publications and source records attributed to Craft, D..

3 recordsLinked to original sources

bioRxiv2020

DNA transduction in Sodalis species: implications for the genetic modification of uncultured endosymbionts of insects

Bacteriophages (phages) are ubiquitous in nature. These viruses play a number of central roles in microbial ecology and evolution by, for instance, promoting horizontal gene transfer (HGT) among bacterial species. The ability of phages to mediate HGT through transduction has been widely exploited as an experimental tool for the genetic study of bacteria. As such, bacteriophage P1 represents a prototypical generalized transducing phage with a broad host range that has been extensively employed in the genetic manipulation of Escherichia coli and a number of other model bacterial species. Here we demonstrate that P1 is capable of infecting, lysogenizing and promoting transduction in members of the bacterial genus Sodalis, including the maternally inherited insect endosymbiont Sodalis glossinidius. While establishing new tools for the genetic study of these bacterial species, our results suggest that P1 may be used to deliver DNA to many Gram negative endosymbionts in their insect host, thereby circumventing a culturing requirement to genetically manipulate these organisms. SummaryA large number of economically important insects maintain intimate associations with maternally inherited endosymbiotic bacteria. Due to the inherit nature of these associations, insect endosymbionts cannot be usually isolated in pure culture nor genetically manipulated. Here we use a broad-host range bacteriophage to deliver exogenous DNA to an insect endosymbiont and a closely related free-living species. Our results suggest that broad host range bacteriophages can be used to genetically alter insect endosymbionts in their insect host and, as a result, bypass a culturing requirement to genetically alter these bacteria.

microbiology↗

bioRxiv2017

Selective and mechanistic sources of recurrent rearrangements across the cancer genome

Cancer cells can acquire profound alterations to the structure of their genomes, including rearrangements that fuse distant DNA breakpoints. We analyze the distribution of somatic rearrangements across the cancer genome, using whole-genome sequencing data from 2,693 tumor-normal pairs. We observe substantial variation in the density of rearrangement breakpoints, with enrichment in open chromatin and sites with high densities of repetitive elements. After accounting for these patterns, we identify significantly recurrent breakpoints (SRBs) at 52 loci, including novel SRBs near BRD4 and AKR1C3. Taking into account both loci fused by a rearrangement, we observe different signatures resembling either single breaks followed by strand invasion or two separate breaks that become joined. Accounting for these signatures, we identify 90 pairs of loci that are significantly recurrently juxtaposed (SRJs). SRJs are primarily tumor-type specific and tend to involve genes with tissue-specific expression. SRJs were frequently associated with disruption of topology-associated domains, juxtaposition of enhancer elements, and increased expression of neighboring genes. Lastly, we find that the power to detect SRJs decreases for short rearrangements, and that reliable detection of all driver SRJs will require whole-genome sequencing data from an order of magnitude more cancer samples than currently available.

cancer biology↗

bioRxiv2016

The value of prior knowledge in machine learning of complex network systems

Our overall goal is to develop machine learning approaches based on genomics and other relevant accessible information for use in predicting how a patient will respond to a given proposed drug or treatment. Given the complexity of this problem, we begin by developing, testing, and analyzing learning methods using data from simulated systems, which allows us access to a known ground truth. We examine the benefits of using prior system knowledge and investigate how learning accuracy depends on various system parameters as well as the amount of training data available. The simulations are based on Boolean networks - directed graphs with 0/1 node states and logical node update rules - which are the simplest computational systems that can mimic the dynamic behavior of cellular systems. Boolean networks can be generated and simulated at scale, have complex yet cyclical dynamics, and as such provide a useful framework for developing machine learning algorithms for modular and hierarchical networks such as biological systems in general and cancer in particular. We demonstrate that utilizing prior knowledge (in the form of network connectivity information), without detailed state equations, greatly increases the power of machine learning algorithms to predict network steady state node values (\"phenotypes\") and perturbation responses (\"drug effects\").

bioinformatics↗