Open Ruiqi-CUB opened 1 year ago
This is the non-symbiotic cockle I use for the control. I just feel like it doesn't make sense if I still keep all the 0s.
@z0on also inquiring if you have code that could be used to remove 0 gene branches as I have a lot of BP and MF terms to cut down the plot. Further, is there also code for the stats portion to output all terms that make a 5% FDR cutoff instead of 10%? Thank you!
@Ruiqi-CUB or if you ever were able to solve this?
Hi folks - look at the “level” options (level1, level2, level3) of the gomwuPlot function. they define significance cutoffs used for plotting.
On Sun, Jul 28, 2024 at 2:46 PM Danielle M. Becker @.***> wrote:
@Ruiqi-CUB https://github.com/Ruiqi-CUB or if you ever were able to solve this?
— Reply to this email directly, view it on GitHub https://github.com/z0on/GO_MWU/issues/16#issuecomment-2254608609, or unsubscribe https://github.com/notifications/unsubscribe-auth/ABZUHGBLLKUUM4O6J5P4BC3ZOU4ADAVCNFSM6AAAAABLS7OAQ2VHI2DSMVQWIX3LMV43OSLTON2WKQ3PNVWWK3TUHMZDENJUGYYDQNRQHE . You are receiving this because you were mentioned.Message ID: @.***>
-- cheers Misha matzlab.weebly.com
“0 gene branches”? I assume those are the terms which have 0/123 or something like that in front? Those are true things, don’t remove them. Zero means that none of the genes within the term pass the arbitrary absValue cutoff, but it does not mean there is no significant shift of the term as a whole.
On Sun, Jul 28, 2024 at 1:48 PM Danielle M. Becker @.***> wrote:
@z0on https://github.com/z0on also inquiring if you have code that could be used to remove 0 gene branches as I have a lot of BP and MF terms to cut down the plot. Further, is there also code for the stats portion to output all terms that make a 5% FDR cutoff instead of 10%? Thank you!
— Reply to this email directly, view it on GitHub https://github.com/z0on/GO_MWU/issues/16#issuecomment-2254593266, or unsubscribe https://github.com/notifications/unsubscribe-auth/ABZUHGEIVQCCKEXDMN6CTJTZOUVF5AVCNFSM6AAAAABLS7OAQ2VHI2DSMVQWIX3LMV43OSLTON2WKQ3PNVWWK3TUHMZDENJUGU4TGMRWGY . You are receiving this because you were mentioned.Message ID: @.***>
-- cheers Misha matzlab.weebly.com
Thank you! I was just seeing if there was anyway to reduce the number of terms shown on the plots as I have too many displayed making the plot very hard to see but I will just play with the dimensions a bit more.
To reduce number of terms displayed I usually just set stricter “levels” cutoffs in the call to gomwuPlot.
Too many terms is a red flag, though. This could happen if you are using ”module-like” measure (0 outside module, kME within module) with simple MWU test, for example. Please make sure you set the analysis correctly!
Misha
On Mon, Jul 29, 2024 at 10:16 AM Danielle M. Becker < @.***> wrote:
Thank you! I was just seeing if there was anyway to reduce the number of terms shown on the plots as I have too many displayed making the plot very hard to see but I will just play with the dimensions a bit more.
— Reply to this email directly, view it on GitHub https://github.com/z0on/GO_MWU/issues/16#issuecomment-2256067491, or unsubscribe https://github.com/notifications/unsubscribe-auth/ABZUHGBHH6JBCHBEIXOS5YTZOZFC3AVCNFSM6AAAAABLS7OAQ2VHI2DSMVQWIX3LMV43OSLTON2WKQ3PNVWWK3TUHMZDENJWGA3DONBZGE . You are receiving this because you were mentioned.Message ID: @.***>
-- cheers Misha matzlab.weebly.com
Sounds good, I will look into this. Thank you for the help!
Hi Misha,
I am using log fold change as the continues measurement of change, and I set the absvalue=1 as the cutoff for plotting. I got quite a lot 0 good genes on the tree. How (or should I do that?) do I cut the branches of those 0 good gene branches? or should I change any parameter to let those branches not inlcuded on the tree? Thanks a lot!
Ruiqi