Members of IFR’s National Collection of Yeast Cultures (NCYC) have joined forces with computer scientists at the University of East Anglia (UEA) to validate novel approaches to constructing a tree of life.
Finding evolutionary trees that best describe how species or sub-species are related to one another is a vital part of NCYC’s research, underpinning our understanding of the collection and how we may exploit it in biological and bioindustrial projects.
Such a task is usually carried out by comparing some aspect of each organism, such as how it looks (morphology) or its genetic code, to those of all the other organisms. The results of these comparisons are then analysed with specialist computer software and a tree is produced. However, UEA’s new methodology, called the “Lasso”, is able to find an evolutionary tree even when some of those comparisons are missing.
The NCYC team, Dr Ian Roberts and Dr Jo Dicks, were delighted to be asked by Dr Katharina Huber and her PhD student George Kettleborough to evaluate Lasso on their datasets. Together the team developed a series of tests designed to put Lasso through its paces. Happily, the software not only managed to find well-established evolutionary patterns in their datasets, even when large numbers of comparisons were deliberately excluded, but it outperformed alternative approaches to the same problem.
The NCYC are leading a major new project to characterise the genetic codes of hundreds of yeast strains within the collection, and plan to use Lasso in establishing the inter-relatedness of the strains.
NCYC is supported as a National Capability by the Biotechnology and Biological Sciences Research Council.
Lasso is a great idea and an important new component within the tree-building tooklit – Dr Jo Dicks, computational biologist within NCYC
“With more of our research being carried out in a high-throughput manner, missing data are ever more inevitable, and approaches such as Lasso ever more needed,” said Dr Ian Roberts, NCYC Curator. “This work represents an important step forward for exploiting the yeast gene pool in commercial processes and highlights once more the value of our collaboration with the UEA”.
Dr Katharina Huber responded “Projects such as the one undertaken by Drs Dicks and Roberts present exciting new challenges for Computer Science and it is encouraging to see that Lasso has the potential to help achieve the goals of their project.”
Reconstructing (super)trees from data sets with missing distances: Not all is lost, George Kettleborough et al, Molecular Biology and Evolution doi: 10.1093/molbev/msv027