alexg9010
commented
1 year ago Hi Xiangyi,
To redefine the grouping and check the difference between the A and (B,C,D,E) groups, you can use the reorganize() function ( see https://www.bioconductor.org/packages/devel/bioc/vignettes/methylKit/inst/doc/methylKit.html#55_reorganize():_reorganizing_samples_and_treatment_vector_within_methylKit_objects ). This allows you to take the samples and specify new groups (e.g. 0 vs 1), then you can run calculateDiffMeth on the new object.
Any given covariance data would be matching the samples, so this should not be a problem.
Best, Alex
Am Mo., 26. Juni 2023 um 13:51 Uhr schrieb DengEr-1993 < @.***>:
Hello there,
A great package which help me solve many problems.
I have a question just as the title said:
I have several groups , for instance, Group A, B, C, D, E and so on.
As we all know we can compare two of them: A vs B or A vs C. It is ok and it works well.
But if I want to explore the difference between A and the other four groups (B, C, D, E), can I unite the rest four groups together as 0 and Group A as 1 to do calculateDiffMeth function ?
On the one hand, I am not sure because the covariance (or specific group characteristic ?) exists between different groups, so I unite B, C, D, E together here will cause uncertain results.
On the other hand, if I can do this as usual, what extra solution should I take to deal with this influence or I just ignore it ?
So I hope you could give me some advice on this question.
I would really appreciated it if you can answer question.
Thank you in advance.
Best, xiangyi
— Reply to this email directly, view it on GitHub https://github.com/al2na/methylKit/issues/289, or unsubscribe https://github.com/notifications/unsubscribe-auth/ADK7JD5VAHYIYMAXN2ZWSU3XNFZVNANCNFSM6AAAAAAZUBOVC4 . You are receiving this because you are subscribed to this thread.Message ID: @.***>
DengEr-1993
Hello there,
A great package which help me solve many problems.
I have a question just as the title said:
I have several groups , for instance, Group A, B, C, D, E and so on.
As we all know we can compare two of them: A vs B or A vs C. It is ok and it works well.
But if I want to explore the difference between A and the other four groups (B, C, D, E), can I
unitethe rest four groups together as 0 and Group A as 1 to docalculateDiffMethfunction ?On the one hand, I am not sure because the covariance (or specific group characteristic ?) exists between different groups, so I unite B, C, D, E together here will cause uncertain results.
On the other hand, if I can do this as usual, what extra solution should I take to deal with this influence or I just ignore it ?
So I hope you could give me some advice on this question.
I would really appreciated it if you can answer question.
Thank you in advance.
Best, xiangyi