Structure-based multiple sequence alignments

Noah Daniels
Massachusetts Institute of Technology
Computer Science and Artificial Intelligence Laboratory

There exist several approaches to improving protein sequence alignment algorithms using structural information. However, how to best balance sequence alignment quality with structural alignment quality has not been clear. We will first demonstrate how sequence information can improve structural alignments, and then explore how advances in homology detection such as Markov random field approaches can improve sequence alignments. Finally, we will discuss how we might re-evaluate the tradeoffs between sequence and structural alignment quality when they are in disagreement.

Presentation (PDF File)

Back to Multiple Sequence Alignment