A short post regarding fig. 4G-H

April 19, 2024

2024   ·   Explanation   Results     ·   Q&A  

Since the publication of our paper, we have received many constructive feedbacks and questions. I decided to provide answers to some of the questions here so more people will be benefit from those important questions.

Some people asked about fig. 4H in the original paper, in particular, what exactly is shown on the x and y axes?

Fig.1 Panel G and H of fig.4 from our original paper.

I have made a diagram to explain fig. 4G and 4H. There are two components:

Fig.2 Diagram to explain results in fig.4G-H. Note, this figure is just to illustrate how we performed the analysis (not real data).

  1. First, we plotted bar graphs for the distribution of replays across trial blocks (panel A and B in the diagram, corresponds to fig. 4G in the paper) for each individual recording session.

  2. Then, we build the scatter plot showing the correlation between maze and post-sleep replay patterns pooled from all data points in all sessions (panel C in the diagram, corresponds to fig. 4H in the paper). To further explain how those two components are related to each other:

  • For each trial block in each example session, we took its replay proportion (#of replay in a particular trial block / total # of replay in that session) during maze as x-axis value and replay proportion during post-sleep as y-axis value of the summary scatter plot (panel C here and fig. 4H in the paper).

    • To give a concrete example, suppose the trial block 5-10 in example session 1 accounts for 100% of the maze replays in that session, and that same trial block also accounts for 100% of the post-sleep replays (panel A), then this data point would locate at the (1,1) coordinate of the summary plot in panel C.

    • The example data point in panel B follows the same logic.

  • Note that proportion values (y-axis values in A and B) are normalized (scaled between 0 and 1) so that they can be compared across sessions.

We had very limited space to explain all our interesting results in detail in the original paper. I hope the explanation here helps!

Further feedbacks in general (including hot takes and criticisms) are welcome! (I love hot takes)! Feel free to leave a comment here or email us (winnieyangwn96@gmail.com, Gyorgy.Buzsaki@nyulangone.org).