Dear chenhcs,
I hope this message finds you well. I am currently studying the FRoGS model with great interest and have learned a lot from your excellent research work.
I have a specific technical question regarding the gene similarity calculation in the model, particularly in the gene_sim_mean_std() function. In the L1000_PhaseI_and_II.csv file, I noticed that each compound perturbation has both UP and DN regulated genes.
Could you kindly clarify the approach you used for calculating gene embedding similarities? Specifically:
- Did you calculate similarities separately for UP and DN genes?
- Or did you treat the UP and DN genes of a compound as a single, integrated set?
Your insights would be invaluable in helping me understand the nuanced design of the gene embedding computation in the FRoGS model.
Thank you for your time and consideration.
Dear chenhcs,
I hope this message finds you well. I am currently studying the FRoGS model with great interest and have learned a lot from your excellent research work.
I have a specific technical question regarding the gene similarity calculation in the model, particularly in the
gene_sim_mean_std()function. In the L1000_PhaseI_and_II.csv file, I noticed that each compound perturbation has both UP and DN regulated genes.Could you kindly clarify the approach you used for calculating gene embedding similarities? Specifically:
Your insights would be invaluable in helping me understand the nuanced design of the gene embedding computation in the FRoGS model.
Thank you for your time and consideration.