{"id":{"repo_id":"mit","oai_identifier":"oai:dspace.mit.edu:1721.1/143287"},"canonical_url":"https://search.dev.ndltd.org/etd/mit/oai:dspace.mit.edu:1721.1/143287","repository":{"repo_id":"mit","name":"MIT","base_url":"https://dspace.mit.edu/oai/request"},"display":{"title":"Deep Neural Networks for Learning Protein Vibrational Behaviors to Characterize Structure and Function","abstract":"Proteins’ structures and motions are essential for nearly all biological functions and malfunctions, making them prime targets for uncovering and controlling processes associated with metabolism and disease. Normal mode analysis is a powerful method that allows us to understand the mechanisms of these functions in high detail, but not without significant cost. Replacing this method with inference by a machine learning model could potentially eliminate this restriction while still providing useful accuracy. Prior work has demonstrated success in a simplified version of this problem that used features computed from each protein’s structure, and predicted parameters for a geometric function-of-best-fit relating the modes, not the explicit modes themselves. In this work, we seek to develop a fully end-toend model that will allow researchers to predict a protein’s normal mode spectrum directly from its peptide sequence, allowing us to bypass the costs associated with both normal mode analysis and protein structure determination. We additionally explore the parallels between protein science and music theory, and provide analysis of a deep neural network trained to understand Bach’s highly structured Goldberg Variations.","abstract_html":"Proteins’ structures and motions are essential for nearly all biological functions and malfunctions, making them prime targets for uncovering and controlling processes associated with metabolism and disease. Normal mode analysis is a powerful method that allows us to understand the mechanisms of these functions in high detail, but not without significant cost. Replacing this method with inference by a machine learning model could potentially eliminate this restriction while still providing useful accuracy. Prior work has demonstrated success in a simplified version of this problem that used features computed from each protein’s structure, and predicted parameters for a geometric function-of-best-fit relating the modes, not the explicit modes themselves. In this work, we seek to develop a fully end-toend model that will allow researchers to predict a protein’s normal mode spectrum directly from its peptide sequence, allowing us to bypass the costs associated with both normal mode analysis and protein structure determination. We additionally explore the parallels between protein science and music theory, and provide analysis of a deep neural network trained to understand Bach’s highly structured Goldberg Variations.","abstract_has_math":false,"creators":["Granberry Jr., Darnell Scott"],"institution":"Massachusetts Institute of Technology","degree_name":"Master","degree_level":null,"degree_discipline":null,"degree_department":"Massachusetts Institute of Technology. Department of Electrical Engineering and Computer Science","school":null,"contributors":[],"advisors":["Buehler, Markus J."],"committee_chairs":[],"committee_members":[],"year":2022,"date_issued":"2022-02","date_published":"2022-02","updated_at":"2026-07-22T22:21:47Z","subjects":[],"languages":[],"rights":["In Copyright - Educational Use Permitted","Copyright MIT"],"rights_urls":["http://rightsstatements.org/page/InC-EDU/1.0/"],"identifier_entries":[]},"links":{"outbound_url":"https://hdl.handle.net/1721.1/143287","outbound_label":"Handle","outbound_source":"dc:identifier.uri"},"metadata_groups":[{"id":"people","label":"People","entries":[{"key":"dc:contributor.advisor","label":"Advisor","values":["Buehler, Markus J."]},{"key":"dc:contributor.department","label":"Department","values":["Massachusetts Institute of Technology. Department of Electrical Engineering and Computer Science"]},{"key":"dc:creator","label":"Author","values":["Granberry Jr., Darnell Scott"]}]},{"id":"academic_context","label":"Academic Context","entries":[{"key":"dc:date.accessioned","label":"Dc Date Accessioned","values":["2022-06-15T13:09:54Z"]},{"key":"dc:date.available","label":"Dc Date Available","values":["2022-06-15T13:09:54Z"]},{"key":"dc:date.issued","label":"Date","values":["2022-02"]},{"key":"dc:publisher","label":"Institution","values":["Massachusetts Institute of Technology"]},{"key":"dc:type","label":"Dc Type","values":["Thesis"]},{"key":"thesis:degree_name","label":"Degree Name","values":["Master","Master of Engineering in Electrical Engineering and Computer Science"]}]},{"id":"language_rights","label":"Language and Rights","entries":[{"key":"dc:rights","label":"Dc Rights","values":["In Copyright - Educational Use Permitted","Copyright MIT"]},{"key":"dc:rights.uri","label":"Rights URI","values":["http://rightsstatements.org/page/InC-EDU/1.0/"]}]},{"id":"identifiers","label":"Identifiers","entries":[{"key":"dc:identifier.uri","label":"Identifier URI","values":["https://hdl.handle.net/1721.1/143287"]}]},{"id":"additional","label":"Additional Metadata","entries":[{"key":"dc:description.abstract","label":"Abstract","values":["Proteins’ structures and motions are essential for nearly all biological functions and malfunctions, making them prime targets for uncovering and controlling processes associated with metabolism and disease. Normal mode analysis is a powerful method that allows us to understand the mechanisms of these functions in high detail, but not without significant cost. Replacing this method with inference by a machine learning model could potentially eliminate this restriction while still providing useful accuracy. Prior work has demonstrated success in a simplified version of this problem that used features computed from each protein’s structure, and predicted parameters for a geometric function-of-best-fit relating the modes, not the explicit modes themselves. In this work, we seek to develop a fully end-toend model that will allow researchers to predict a protein’s normal mode spectrum directly from its peptide sequence, allowing us to bypass the costs associated with both normal mode analysis and protein structure determination. We additionally explore the parallels between protein science and music theory, and provide analysis of a deep neural network trained to understand Bach’s highly structured Goldberg Variations."]},{"key":"dc:description.degree","label":"Dc Description Degree","values":["M.Eng."]},{"key":"dc:title","label":"Title","values":["Deep Neural Networks for Learning Protein Vibrational Behaviors to Characterize Structure and Function"]}]}],"canonical_facts":{"dc:contributor.advisor":["Buehler, Markus J."],"dc:contributor.department":["Massachusetts Institute of Technology. Department of Electrical Engineering and Computer Science"],"dc:creator":["Granberry Jr., Darnell Scott"],"dc:date.accessioned":["2022-06-15T13:09:54Z"],"dc:date.available":["2022-06-15T13:09:54Z"],"dc:date.issued":["2022-02"],"dc:description.abstract":["Proteins’ structures and motions are essential for nearly all biological functions and malfunctions, making them prime targets for uncovering and controlling processes associated with metabolism and disease. Normal mode analysis is a powerful method that allows us to understand the mechanisms of these functions in high detail, but not without significant cost. Replacing this method with inference by a machine learning model could potentially eliminate this restriction while still providing useful accuracy. Prior work has demonstrated success in a simplified version of this problem that used features computed from each protein’s structure, and predicted parameters for a geometric function-of-best-fit relating the modes, not the explicit modes themselves. In this work, we seek to develop a fully end-toend model that will allow researchers to predict a protein’s normal mode spectrum directly from its peptide sequence, allowing us to bypass the costs associated with both normal mode analysis and protein structure determination. We additionally explore the parallels between protein science and music theory, and provide analysis of a deep neural network trained to understand Bach’s highly structured Goldberg Variations."],"dc:description.degree":["M.Eng."],"dc:identifier.uri":["https://hdl.handle.net/1721.1/143287"],"dc:publisher":["Massachusetts Institute of Technology"],"dc:rights":["In Copyright - Educational Use Permitted","Copyright MIT"],"dc:rights.uri":["http://rightsstatements.org/page/InC-EDU/1.0/"],"dc:title":["Deep Neural Networks for Learning Protein Vibrational Behaviors to Characterize Structure and Function"],"dc:type":["Thesis"],"thesis:degree_name":["Master","Master of Engineering in Electrical Engineering and Computer Science"]},"updated_at":"2026-07-22T22:21:47Z"}