Carnegie Mellon University
February 05, 2018

Lane Fellow DeBlasio Publishes Book

Dan DeBlasioLane Fellow Dan DeBlasio, along with his PhD advisor John Kececioglu, recently published a book titled "Parameter Advising for Multiple Sequence Alignment”, which details a method to select input-specific parameter
choices for the fundamental problem of multiple sequence alignment.

Using Parameter Advising can produce higher quality alignments, which leads to improvements in downstream analyses like protein structure prediction, phylogenetic inference, or homology search. The accuracy of multiple
sequence alignments can be greatly impacted by choosing non-ideal parameter settings (and in turn the downstream analyses), and while the default parameter settings work well on average, they may produce poor results on some input. The book also describes Facet (short for Feature-based Accuracy
Estimator) which when used with Parameter Advising can automatically select improved parameters for any input set of protein sequences.

“Parameter Advising for Multiple Sequence Alignment” is part of Springer International’s Computational Biology Series.