Hi! I'm Yijie, a.k.a. 银怡杰

About Me

I'm a PhD student / Research Assistant at the University of Cambridge, supervised by Prof Albert Cardona. I am passionate about interpreting the comprehensive neural wiring diagram (the 'connectome') of the fruit fly brain. This is because:

• I appreciate things that are efficient and intelligent.
• The brain is not only efficient, but also 'embodies intelligence': neurons are quite similar to each other. Just by arranging the elements in a smart way, you get a highly capable and flexible system. Wouldn't it be great if we could learn from this?
• Something like a connectome seems essential to understand the brain without having to wave my hands.
• So based on my definition of understanding and the current size of connectomes, it seems necessary to start small.

I hope to understand the 'algorithms of information processing in the brain', and perhaps later apply those in designing artificial systems, to not only check our understanding, but also to learn from nature.

For leisure, I like singing, playing basketball, roller blading and ice skating. I also enjoy good food.

My Projects

• [ongoing] Connectome Interpreter: I realised that many operations in connectome analysis are repetitive, time-consuming, and can be automated. I also realised that some Explainable AI tools fit almost perfectly for connectome analysis. So both are included in this package.
• [ongoing] I am comparing the learning & memory circuit of the adult, larval fruit fly, and the megaphragma.
• [ongoing] I am working to determine the neurotransmitter expression profile of all neurons in the fly larva.
• I contributed significantly to the first full-brain connectome of an adult fly, published here and here:
• As a Research Assistant at the Drosophila Connectomics Group, Department of Zoology, University of Cambridge, I reconstructed many neurons (neuron reconstruction is a process of manually fixing errors of image segmentation, in this case the images are electron microscopy images of a fly brain; and segments are neurons.). This gave me a good understanding of the electron microscopy images, the reconstruction process, and the errors the model might make.
• I contributed significantly to the hemilineage (the developmental origin of neurons) annotations of neurons. This project gave me a very good understanding of neuroanatomy across the brain.
• I annotated cell types for the Kenyon Cell (KC) subtypes in this dataset, and compared cell numbers and connectivity with the hemibrain dataset. We showed that there's a two-fold cell number difference in the KCg-m subtype across individuals. We find that when there are more cells, each one essentially "take less responsibility", and receive less from both excitatory and inhibitory inputs.
• In Eckstein, Bates et al. 2024, I showed that based on the predicted neurotransmitters, there might have been a developmental switch of neurotransmitter expression in one of the hemilineages.

Contact

You can reach me at: yy432-at-cam.ac.uk or through LinkedIn