How to understand the `weight` and `p_value` in the result of get_progeny.

[xiangpin]

Hello Thanks for developing the packages. I used it to infer the pathway activity in my data. But I don't understand the weight and p_value in the result of get_progeny from PROGENy.

• What do the negative weight and positive weight values mean respectively?

• What does the p_value mean? Can we select the targets based on the p_value?

[PauBadiaM]

Hi @xiangpin

The weights indicate how each gene is associated to a given pathway. A positive Pathway-Gene weight indicates that that gene was seen to be upregulated in stimulation experiments for that pathway. A negative weight indicates that that gene was seen to be downregulated when that pathway was stimulated. The p-value tells how confident are we about these weights. You can find the exact protocol on how the weights and p-values where extracted from a collection of stimulation experiments in the original publication of progeny: https://doi.org/10.1038/s41467-017-02391-6. Regarding the p-values, we use them to rank Pathway-Gene interactions and then we select the top N, N being an arbitrary integer number (which corresponds to the top argument of the get_progeny function). The smaller N is, the more confident we are that those genes are associated to a given pathway but we have low coverage, with bigger N we have the opposite effect, larger coverage but less confidence. Hope this was helpful!