University of Montana computer science researcher Travis Wheeler recently was awarded a four-year $1.1 million grant to develop machine learning approaches to improving the accuracy and speed of sequence annotation.

Sequence annotation involves identifying and labeling features in DNA, RNA and protein sequences. It’s a critical step for helping understand the evolution, function and patterns of activity for these biological components.

While this is his first National Institutes of Health Research Project Grant, Wheeler is no stranger to securing funding. He is in his final year of investigator funding from UM’s Center for Biomolecular Structure and Dynamics, and his research group also is supported by two other NIH grants.

“Modern molecular biology depends on effective methods for annotating biological sequences quickly and accurately,” Dr. Wheeler said. “Mutations and sequencing errors are sort of like misspellings in biological sequences, so our software needs to be able to annotate sequences despite misspellings – while at the same time avoiding incorrect annotations.”

He said the scale of genome sequencing efforts today is breathtaking, and this requires the development of increasingly fast ways to annotate those sequences.

“To reach these goals, students working with me will build on recent advances in a field called deep learning, which is at the heart of modern approaches to text processing and image labeling,” Wheeler said. “I’m excited to get started!”