And in natural mycospic because of its superior spatial resolution, one can look at a vesicle and density the PSD and one can [COUGH] trace this synapse if one has a different section of a different brain. So this illustrates segmentation that people can label based on electrical micrograph which cell is which cell. And then, if you imagine, that if you have lost tracing, you can trace which cell is which cell. Then, if you cut different depths of this. And in reconstruction you can achieve with high spatial resolution the kinetomics of a mouse brain or part of the human brain, okay? Again, there are increasing efforts for different groups are trying to do that. Right now the difficulty is how to cut perfect slice. And how to perfectly align them. How to computationally, efficiently trace them without the human intervention. Because traditionally people can only, by eye and each person can trace and that takes a lot of efforts, including some of my students in the lab. So how do you allow it using artificial intelligence? Some smart computer algorithm to automatically trace them and then reconstruct them. Those will pose challenges for computer scientists, for students whose major or second major is in math to develop techniques for those reconstructions. Again, this is a one requires real interdisciplinary approaches. Because, I guess 90% of the computer scientists doesn't care or didn't know there is this huge challenges in brain reconstruction, okay? And 80% of the brain scientists doesn't know how to use the computer science or advanced algorithm techniques to trace those brain regions. And even for 20% of them I know those techniques, 90% of them doesn't know how to compute the electron micrograph to cloud the status. They don't know how to use a diamond knife to cut the brain slice. Okay, so it really requires the interdiscipline approaches. And we can go on, and on, and on. And this illustrates that one reason develop techniques again using the virus to trace the neuron connections. We just mentioned that the virus has this problem. If we infect the cells, it will just reproduce and then spread, label the whole bread. So how do use a virus to label specific cell, or a cell assembly that are connected? Well, Edward Holloway and his collaborators designed a mass strategy. Furthermore, how do you label a specific cell? Well, they engineered a virus because virus requires using a envelope protein to target a cell to its [INAUDIBLE]. Okay, so they engineer the virus, only put in a virus with a special envelope protein. And this envelope protein will only recognize this specific TVA receptor. For example, this TVA receptor is from a chicken virus. Okay, and if you only it's present in mouse. There's no chicken receptor there, okay? So only is present this receptor in a cell of interest to, you are interested in. Then you are putting this virus, then this virus can only infect this cell, okay? How do you only express this TVA in this cell? Well, we just talked about you can use this tissue specific promoter, okay? Express this TVA in itself, okay? And then, this virus can only express in this cell. And then, this cell was different color, for example, it was GIP, okay? But then what you can do is that this virus can only reproduce in your cell, okay? Because you put in this virus lacking a special gene. So it cannot self reproduce unless it is present in this cell. Where in this cell you also express a special gene that this virus is lacking, okay? So, only when this virus is pressed in this cell. And then the cell also provide a glycoprotein to compliment this virus and then this virus can reproduce, get release, okay? And once it release, then it can retrogradely label other cells that has this synaptic connection with it, okay? And once it label this cell because of the other cell do not have this glycoprotein. Then those virus in that other cell will not reproduce, okay? So, you only label this cell, okay? So essentially, you allowed this virus to spread only one step, okay? And you're only looking at one cell and its connecting cell and no other cell, okay. And using this method, people have generating some specific connection map for certain cells. And then you can imagine this vessel can be reused again. For example, if you're using and another color and yet another complementation in these cells. And then you can from this cell to label another cell with a different color. Okay, so you can use over and over again to trace the one synapse connection and two synapse connection.