Two of our undergraduate students received Magellan Scholar awards. Logan Hood received an award for his project on "Affordable Medical Ultrasound Training Simulator" and Garret DeBruin for his project titled "Classification and Detection of Proteotypic Peptides Through The Use of Hidden Markov Models and Neural Networks." Congratulations to both of them! Garret DeBruin's project abstract:

Classification and Detection of Proteotypic Peptides Through The Use of Hidden Markov Models and Neural Networks In traditional peptide sequencing an unknown peptide’s MSMS spectrum is used as a query to a database. This query yields a set of peptides with spectra similar to the observed spectrum. A more difficult problem is De Novo peptide sequencing, which does not depend on a database of known peptides. De Novo peptide sequencing has the problem of considering all possible peptides consistent with the observed spectrum, and ranking them according to some scoring function. Proteotypic peptides are peptides that are easily fragmented by collision induced dissociation and are easily detected by tandem mass spectrometers. In contrast, nonproteotypic peptides are unlikely to be detected. Under the proposed approach, it would be possible to improve de novo sequencing results by discounting nonproeotypic peptides since they are unlikely to be detected even though they are consistent with the observed spectrum. This would significantly improve the accuracy of the sequencing.