EduEyras
commented
1 year ago Hi,
thanks for your email
I’d assume you run StringTie2 in a way to obtain a single transcriptome for all sample, and one TPM per transcript per sample, is that correct?
And then each condition have 2 or more samples?
Do you have enough TPM values for each transcript across samples? Or do you see many missing values? That could be one of the problems
You can change the proportion of missing values allowed with the -na parameter. But it will work only if you have enough samples per condition.
I hope this helps
Eduardo
On Sat, 22 Jul 2023 at 2:48 pm, Renee-SU @.***> wrote:
Hello SUPPA developers,
As limited number of AS events were detected with known annotation file, I employed StringTie to assemble and merge a gtf file. Then I used this new gtf for following steps.
When I used the command suppa.py diffSplice --method empirical --input all.ioe --psi 1R_Mk.psi 1R_Fp.psi --tpm 1R_Mk.tpm 1R_Fp.tpm -o 1R_Mk_vs_Fp --lower-bound 0.1 -gc
The first page of dpsi output file is like this: all rows of the dpsi column were nan and all rows of p value were 1.0
[image: 1690000365733] https://user-images.githubusercontent.com/139733263/255311363-61e6bea3-ca26-4b58-8855-957464c667b4.png Then, I modified the --method parameter with classical and here is the command suppa.py diffSplice --method classical --input all.ioe --psi 1R_Mk.psi 1R_Fp.psi --tpm 1R_Mk.tpm 1R_Fp.tpm -o 1R_Mk_vs_Fp --lower-bound 0.1
The fist page of dpsi output file is like this: not all rows of the dpsi column were nan and except for 1.0, other values showed in the p value column. [image: 1690000779529] https://user-images.githubusercontent.com/139733263/255311586-47714698-d047-4ef4-9b64-df08fd354a64.png However, when I filtered the p value column with the condition "p < 0.05", none was left. All rows of p value are greater than 0.05.
Have you encountered such problem before? Any possible reasons?
If this problem cannot be solved or fixed, could I only apply the condition dpsi > 0.1 not p < 0.05?
Cheers, Renee
— Reply to this email directly, view it on GitHub https://github.com/comprna/SUPPA/issues/166, or unsubscribe https://github.com/notifications/unsubscribe-auth/ADCZKBZABXG3GHQWIY5IW5DXRNLSNANCNFSM6AAAAAA2TTARXA . You are receiving this because you are subscribed to this thread.Message ID: @.***>
--
Renee-SU
Hello SUPPA developers,
As limited number of AS events were detected with known annotation file, I employed StringTie to assemble and merge a gtf file. Then I used this new gtf for following steps.
When I used the command suppa.py diffSplice --method empirical --input all.ioe --psi 1R_Mk.psi 1R_Fp.psi --tpm 1R_Mk.tpm 1R_Fp.tpm -o 1R_Mk_vs_Fp --lower-bound 0.1 -gc
The first page of dpsi output file is like this: all rows of the dpsi column were nan and all rows of p value were 1.0
The fist page of dpsi output file is like this: not all rows of the dpsi column were nan and except for 1.0, other values showed in the p value column.
However, when I filtered the p value column with the condition "p < 0.05", none was left. All rows of p value are greater than 0.05.
Have you encountered such problem before? Any possible reasons?
If this problem cannot be solved or fixed, could I only apply the condition dpsi > 0.1 not p < 0.05?
Cheers, Renee