Winter 2019 Final Exam Schedule


Final Examination Schedule


Winter 2019

Thursday, March 7

Todor K. Avramov

Chair: Dr. Dong Si
Candidate: Master of Science in Computer Science & Software Engineering
11:00 A.M.; DISC 464
Deep Learning for Validating Resolution and Detecting Secondary Structure Elements of Proteins in 3D Cryo-Electron Microscopy Images

Cryo-electron microscopy (cryo-EM) is becoming the imaging method of choice for determin­ing protein structures. Many atomic structures have been resolved based on an exponentially growing number of published three-dimensional (3D) cryo-EM density maps. However, the resolution value claimed for the reconstructed 3D density map has been the topic of scien­tific debate for many years. The Fourier Shell Correlation (FSC) is the currently accepted cryo-EM resolution measure, but it can be subjective, manipulated, and has its own lim­itations. This thesis proposes supervised deep learning methods to extract representative 3D features at high, medium and low resolutions from simulated protein density maps and build classification models that objectively validate resolutions of experimental 3D cryo-EM maps. Specifically, classification models based on dense artificial neural network (DNN) and 3D convolutional neural network (3D CNN) architectures are presented. The trained mod­els can classify a given 3D cryo-EM density map into one of three resolution levels: high, medium, low. The DNN model achieved 92.73% accuracy and the 3D CNN model achieved 99. 75% accuracy on simulated test maps. When tested on simulated maps at gradually vary­ing resolutions, the two models identified the resolution boundaries between high, medium and low resolutions. The deep learning models clustered maps lower than 4-4.5Å in the high resolution class, maps between 5.0-8.5Å in the medium resolution, and maps at resolutions>=9.0Å were classified as low resolution. Applying the DNN and 3D CNN models to thirty experimental cryo-EM maps achieved an agreement of 60.0% and 56.7%, respectively, with the author published resolution value of the density maps. These results suggest that deep learning can potentially improve the resolution evaluation process of cryo-EM maps but further work is needed to account for local variability of resolution as suggested by recent studies.

Detection of protein secondary structure elements (SSEs) to aid in the creation of ac­curate atomic models especially from medium resolution cryo-EM maps is another area of current research. Medium resolution experimental cryo-EM images lack detail and contain noise, and thus require additional computational and visualization analyses to fully deter­mine protein structures. Most previous researches proposed prescriptive image-processing and pattern matching algorithms to locate alpha-helices and beta-sheets in cryo-EM maps, but these methods were not fully automated and required subjective selection of parameters. This the­sis explores a convolutional neural network model for end-to-end voxelwise segmentation of 3D cryo-EM density images. The 3D segmentation model, adapted from the U-Net archi­tecture, was constructed in TensorFlow and it optimized a multi-class Tversky loss function with Adam optimization algorithm. The proposed 3D U-Net model was trained to segment a cryo-EM map by classifying each voxel as either being part of an alpha-helix, a beta-sheet, a tum/loop, or background. For that purpose, I first introduce and describe a novel method to generate large amounts of labeled cryo-EM maps suitable for training deep learning models for secondary structure segmentation. The model achieved higher per-class and overall preci­sion and recall rates than previous methods when tested on 3597 simulated cryo-EM density maps. The proposed method was also shown to reliably segment experimental cryo-EM maps. Finally, the 3D U-Net segmentation model was compiled into an executable program and integrated as a plug-in in the UCSF Chimera visualization and analysis system.

Monday, March 11

Supriya Somenahalli

Chair: Dr. William Erdly
Candidate: Master of Science in Computer Science & Software Engineering
11:00 A.M.; DISC 464
Access Control, Autocomplete and Report Generator

The EYE (Educating Young Eyes) Center for Children's Vision Learning & Technology is a University of Washington, Bothell sponsored non-profit organization dedicated to the research, development, and education of technologies to help increase awareness of the importance of functional vision in children learning process. As a part of the EYE Center, many applications have been developed to increase the awareness amongst children. One such application being developed is the Vision Quick check.

The aim of this Vision Quick check application is to promote and popularize the vision tests, especially among children and youth. It contains tests to detect Near Vision problems. To pass the test, the child is required to score beyond a certain score, if not, the child will be referred to a doctor for further inspection. The first version of the application is being targeted for Android phone users.

This project starts by describing how the access control is provided for the Vision Quickcheck application. Since medical data and sensitive student information is involved in the application, access-control plays a very important role. For a nurse to conduct the tests on a child, the nurse would have to be registered prior as a valid tester and the child who would be taking the test would have to be registered as well.

The project then talks about the backend implementation being carried out while selecting students on the phone before conducting the tests. Since the application is designed for school children and would be conducted in schools, the children are registered to the application through the school administrator. To be able to conduct the tests on students, the nurse would require an automated manner to select the students instead of having to enter all the details. Hence, this project also aims in describing the implementation of the student selection property in the application.

Finally, the project also talks about the reports generator in another application called the Centralized web portal. As mentioned earlier, the nurse and the children would need to be registered to the application before the tests are conducted. The registration is done with the help the Centralized Web Portal. The centralized web portal also contains research capabilities. To perform research, report generation is a vital component. The detailed implementation of the report generator is described in this report.

Wednesday, March 13

Christina Burnett

Chair: Dr. Hazeline Asuncion
Candidate: Master of Science in Computer Science & Software Engineering
1:15 P.M.; DISC 464
Automated Assessment of Swagger for Compliance to Standards

Our project is a case study of the automated assessment of application programming interface (API) design quality at T-Mobile. We describe a means to deliver actionable feedback on the designs of HTTP services that aspire to be RESTful. We have shown how a set of API standards for Swagger documents can be translated into programmatically enforceable rules, and how adherence to those rules can be measured and socialized in a corporate environment. The T-Mobile API standards are based on a set of eight API patterns that are derived from the OpenAPI specification. 

Thursday, March 14

Spencer Moritz

Chair: Dr. Dong Si
Candidate: Master of Science in Computer Science & Software Engineering
11:00 A.M.; DISC 464
Deep Learning to Predict Protein Backbone Structure from High-Resolution Cryo-EM Density Maps

Understanding a protein’s structure can lead to the discovery of therapeutic protein drugs. However, imaging proteins remain a challenge due to the small size. Cryo-electron microscopy (cryo-EM) is a leading imaging technology that has recently been able to produce near atomic resolution images called electron density maps. However, predicting various protein structures remains a challenge on all but the most pristine density maps (< 2.5Å resolution). Here we introduce a deep learning model that uses a set of cascaded convolutional neural networks (CNNs) to predict Cα atoms along a protein’s backbone structure. The cascaded-CNN (C-CNN) is a novel deep learning architecture comprised of multiple CNNs, each predicting a specific aspect of a protein’s structure. This model predicts various levels of protein structures: secondary structure elements (SSEs), backbone structure, and Cα atoms. It combines the results of each to produce a complete prediction map. The cascaded-CNN is a sematic segmentation image classifier and was trained using thousands of simulated density maps. This method is largely automatic and only requires a recommended threshold value for each evaluated protein. A specialized tabu-search path walking algorithm was used to produce an initial backbone trace with Cα placements. This method was tested on 50 experimental maps between 2.6Å and 4.4Å resolution. It outperformed several state-of-the-art prediction methods including RosettaES, MAINMAST, and a Phenix based method by producing the most complete prediction models, as measured by percentage of found Cα atoms. This method accurately predicted 88.5% (mean) of the Cα atoms within 3Å of a protein’s backbone structure surpassing the 66.8% mark achieved by the leading alternate method (Phenix based fully automatic method) on the same set of density maps. The C-CNN also achieved an average RMSD of 1.23Å for all 50 experimental density maps which is similar to the Phenix based fully automatic method.

Back to top