Home

Updates From Graduate School

After my last layoff, I decided it was time to change my plan. I’ve reinvented myself professionally a few times, most notably, going from special education teacher to software engineer. Rather than opting for a complete career change, I decided to pivot to the promising field of biological data science.

ASU offers a graduate certification program in biological data science, and based on a few different sources, is expected to be both resistant to AI automation, and is in high demand. So far it’s been fun, and I’m learning a lot. I figured I’d take time to do a write up of each class, and talk about what we’ve learned.

LSC 555, or Integrative Biology was the last class I took. It felt like the first class where we really had to dive head first into the scientific literature. Each week we had to read scientific publications and write up a short summary and answer questions about it. Believe me when I say I was intimidated by the paper Cryo-EM structures of tau filaments
from Alzheimer’s Disease by Fitzpatrick and the gang. Here is my summary:

How are the structures presented in the article related to Alzheimers pathology?  

The structures have been associated with Alzheimer’s disease, and these tau filaments can seed tau aggregation when transplanted into lab animals such as mice. Aggregation of tau filaments has been shown to cause neurodegeneration and dementia. Furthermore, short filaments are the main species of seed-competent tau in modified mice showing the relevant tau mutation.

What techniques were used to determine the structures described in the article? 

The authors used samples of the cerebral cortex from a patient who had been diagnosed neuropathologically with Alzheimer’s, and had lived with it for 10 years. They sequenced her genotype and found no disease-causing mutations, but using Thioflavin S staining, they found a lot of neurofibrillary tangles and neuritic plaques. These are associated with Alzheimer’s disease. The authors were able to separate the PHF and SF structures in a centrifuge, and produced morphologically distinct structures. Using cryo-EM imaging, and software, they were able to use helical reconstruction to create a 3-D model of the structures, showing density and direction of separation of beta-strands, along the helical axis. This allowed the researchers to propose refined atomic models for the cores of PHFs and SFs.

What ways, if any, was data science utilized in this experiment (large data sets)?

Data science was helpful in this experiment by providing a statistical basis for associating tau filaments, tangles, and plaques with a variety of diseases, not just Alzheimer’s. In addition, it also is useful in the software RELION; software that is able to make predictions and reconstruct a 3D model of these microscopic structures. Researchers are able to take a bunch of pictures using cryo-EM imaging and RELION is able to visualize what the helical structure looks like.