palantir.plot.plot_palantir_results

[yitengfei120011]

hi，palantir_entropy is all 0 in my data. Is there anything wrong with my data? All of them are zero， how do we explain that?

[katosh]

Hi @yitengfei120011, thanks for reaching out. From the plot, it appears that only a single terminal state was specified. As a result, the fate probability remains constant at 1 for that fate across the entire space. Consequently, the entropy also remains constant, calculated as:

$E=-\sum_i p_i\cdot log(p_i)=-1\cdot log(1) = 0$

I hope this clarifies the issue. Let me know if you have any further questions!

Edit: In other words, the entropy only comes from the potential of the cell to differentiate into distinct terminal cell types. If multiple terminal cell types are available, then this potential goes down as the probability for a fate approaches 1. If, only one fate is available, then the probability for this fate is constantly 1 and trhe potential remains constantly low.