search

churchlab / AAV_fitness_landscape

Analysis for "Comprehensive fitness landscape of AAV capsid reveals a viral gene and enables machine-guided design"
20 stars 13 forks source link

A20 antibody neutralization assay is not correctly designed #3

Open vuhongai opened 2 years ago

vuhongai commented 2 years ago

Dear authors,

Thank you for awesome work and sharing it in very detailed manner.

I however think that the A20 antibody neutralization assay has an issue with the calculation of antibody selection. The selection in the paper is s = fab/fv but your calculation did not account for the variability of infectivity of different mutants. If you normalize with fv (frequency of the viral packaging), you are assuming that every mutant infects HEK293 at the same rate, which is absolutely not the case, especially with the mutation in the variable regions. So I would propose this selection formula instead: s = f_withA20 / f_withoutA20 where f is the frequency found in the cells after infection.

I hope this might contribute somehow to your work. Best regards, Ai

kelsic commented 2 years ago

Thanks Ai for your suggestion! We agree what you propose seems to be a reasonable approach and an improvement on our initial calculation, provided that the variants which do not infect at all are somehow masked out. -Eric

On Tue, Apr 19, 2022 at 8:54 AM Ai Vu Hong @.***> wrote:

Dear authors,

Thank you for awesome work and sharing it in very detailed manner.

I however think that the A20 antibody neutralization assay has an issue with the calculation of antibody selection. The selection in the paper is s = fab/fv but your calculation did not account for the variability of infectivity of different mutants. If you normalize with fv (frequency of the viral packaging), you are assuming that every mutant infects HEK293 at the same rate, which is absolutely not the case, especially with the mutation in the variable regions. So I would propose this selection formula instead: s = f_withA20 / f_withoutA20 where f is the frequency found in the cells after infection.

I hope this might contribute somehow to your work. Best regards, Ai

— Reply to this email directly, view it on GitHub https://github.com/churchlab/AAV_fitness_landscape/issues/3, or unsubscribe https://github.com/notifications/unsubscribe-auth/AAPL5YKXYSRIYL7YAIJQB5TVF2UHPANCNFSM5TYYBV4A . You are receiving this because you are subscribed to this thread.Message ID: @.***>