search

wososa / PSI-Sigma

PSI-Sigma
Other
35 stars 10 forks source link

Event type #53

Closed ktruve closed 11 months ago

ktruve commented 1 year ago

Hi, We ran PSI-sigma on Danio_rerio using a gtf -file from Ensembl (Danio_rerio.GRCz11.110.chr.sorted.gtf). We are confused by the Event Type that says 'R' or 'S' or 'W'. What does this mean?

(Running the same commands on human data gave the Events Types that we could understand: A3SS A5SS IR IR (overlapping region) MES SES TSS|A3SS TSS|A5SS) Hope you can help us explain the Danio Rerio output. Best, Katarina

wososa commented 1 year ago

Hi @ktruve ,

R , S, and W stand for "Intron Retention", "Alternative splice side", and "exon-inclusion/skipping" events. In PSI-Sigma's output, there should be a sorted.txt file giving you the Event Types that you could understand. Let me know if the file wasn't generated.

Thanks, Woody

ktruve commented 1 year ago

Thank you very much Woody! That was good to know, but no we don't get a file sorted.txt (?) We look in the file PSIsigma_r10_ir3.txt Best, Katarina

wososa commented 1 year ago

Hi Katarina,

Could you please elaborate a bit further? Do you mean that you get the ‘sorted.txt’ file when you use human files but not getting the file when you use zebrafish files?

Could you please share the system message when you are running PSI-Sigma?

Thanks, Woody

On Tue, Oct 31, 2023 at 6:39 AM ktruve @.***> wrote:

Thank you very much Woody! That was good to know, but no we don't get a file sorted.txt (?) We look in the file PSIsigma_r10_ir3.txt Best, Katarina

— Reply to this email directly, view it on GitHub https://github.com/wososa/PSI-Sigma/issues/53#issuecomment-1786952955, or unsubscribe https://github.com/notifications/unsubscribe-auth/ABWLHWOV6TOIL6X3B65WG23YCDIPDAVCNFSM6AAAAAA6WTLMGWVHI2DSMVQWIX3LMV43OSLTON2WKQ3PNVWWK3TUHMYTOOBWHE2TEOJVGU . You are receiving this because you commented.Message ID: @.***>

ktruve commented 1 year ago

Hi again,

yes that is right. We get the sorted file for human. I see now that we get an error message at the end of the run with Danio_rerio: Reading... PSIsigma.db (ERROR) 153345 can't find wings in the database. Aborting... ===Filtering spent 0.0000 hours.=== Archiving... PSIsigma_r10_ir3.txt PSIsigma_r10_ir3.txt

What does that mean? Best,

Katarina

wososa commented 1 year ago

Hi Katarina,

This means what you might be using the wrong PSIsigma.db file. Could you please try running PSI-Sigma for Danio_rerio in a separate folder so that it won’t use human files such as the .db file?

Best, Woody

On Tue, Oct 31, 2023 at 6:54 AM ktruve @.***> wrote:

Hi again,

yes that is right. We get the sorted file for human. I see now that we get an error message at the end of the run with Danio_rerio: Reading... PSIsigma.db (ERROR) 153345 can't find wings in the database. Aborting... ===Filtering spent 0.0000 hours.=== Archiving... PSIsigma_r10_ir3.txt PSIsigma_r10_ir3.txt

What does that mean? Best,

Katarina

— Reply to this email directly, view it on GitHub https://github.com/wososa/PSI-Sigma/issues/53#issuecomment-1786975171, or unsubscribe https://github.com/notifications/unsubscribe-auth/ABWLHWKQHKVRRJAAXZE2PXTYCDKETAVCNFSM6AAAAAA6WTLMGWVHI2DSMVQWIX3LMV43OSLTON2WKQ3PNVWWK3TUHMYTOOBWHE3TKMJXGE . You are receiving this because you commented.Message ID: @.***>

ktruve commented 1 year ago

Hi,

We have Danio_rerio in a separate folder so that could not be the problem. Also deleted everything and reran, but same problem.

Any other suggestions?

Best,

Katarina

Statistics option = Student's t-test

number of p-value = 13378

Number of final p-value = 13121

Skipping p-value adjustment.

number of fdr(BH) = 13121

===PSI analysis spent 0.0119 hours.===

Filtering ΔPSI results...

Filtering mode = 0

Reading... /medstore/databases/research/Danio_rerio/Ensembl/Danio_rerio.GRCz11.110.chr.sorted.gtf.mapping.txt

Reading... PSIsigma.db

(ERROR) 153345 can't find wings in the database.

Aborting...

===Filtering spent 0.0000 hours.===

Archiving... PSIsigma_r10_ir3.txt

PSIsigma_r10_ir3.txt

On 31 Oct 2023, at 11:57, Woody Lin @.***> wrote:

Hi Katarina,

This means what you might be using the wrong PSIsigma.db file. Could you please try running PSI-Sigma for Danio_rerio in a separate folder so that it won’t use human files such as the .db file?

Best, Woody

On Tue, Oct 31, 2023 at 6:54 AM ktruve @.***> wrote:

Hi again,

yes that is right. We get the sorted file for human. I see now that we get an error message at the end of the run with Danio_rerio: Reading... PSIsigma.db (ERROR) 153345 can't find wings in the database. Aborting... ===Filtering spent 0.0000 hours.=== Archiving... PSIsigma_r10_ir3.txt PSIsigma_r10_ir3.txt

What does that mean? Best,

Katarina

— Reply to this email directly, view it on GitHub https://github.com/wososa/PSI-Sigma/issues/53#issuecomment-1786975171, or unsubscribe https://github.com/notifications/unsubscribe-auth/ABWLHWKQHKVRRJAAXZE2PXTYCDKETAVCNFSM6AAAAAA6WTLMGWVHI2DSMVQWIX3LMV43OSLTON2WKQ3PNVWWK3TUHMYTOOBWHE3TKMJXGE . You are receiving this because you commented.Message ID: @.***>

— Reply to this email directly, view it on GitHubhttps://github.com/wososa/PSI-Sigma/issues/53#issuecomment-1786981113, or unsubscribehttps://github.com/notifications/unsubscribe-auth/A5JMZXXMRSCNKD5VLTPDL4LYCDKTHAVCNFSM6AAAAAA6WTLMGWVHI2DSMVQWIX3LMV43OSLTON2WKQ3PNVWWK3TUHMYTOOBWHE4DCMJRGM. You are receiving this because you were mentioned.Message ID: @.***>

wososa commented 1 year ago

Hi Katarina,

Could you delete the PSIsigma.db and run again? It is likely that the .db file wasn’t generated correctly.

Thanks, Woody

On Tue, Oct 31, 2023 at 10:14 AM ktruve @.***> wrote:

Hi,

We have Danio_rerio in a separate folder so that could not be the problem. Also deleted everything and reran, but same problem.

Any other suggestions?

Best,

Katarina

Statistics option = Student's t-test

number of p-value = 13378

Number of final p-value = 13121

Skipping p-value adjustment.

number of fdr(BH) = 13121

===PSI analysis spent 0.0119 hours.===

Filtering ΔPSI results...

Filtering mode = 0

Reading... /medstore/databases/research/Danio_rerio/Ensembl/Danio_rerio.GRCz11.110.chr.sorted.gtf.mapping.txt

Reading... PSIsigma.db

(ERROR) 153345 can't find wings in the database.

Aborting...

===Filtering spent 0.0000 hours.===

Archiving... PSIsigma_r10_ir3.txt

PSIsigma_r10_ir3.txt

On 31 Oct 2023, at 11:57, Woody Lin @.***> wrote:

Hi Katarina,

This means what you might be using the wrong PSIsigma.db file. Could you please try running PSI-Sigma for Danio_rerio in a separate folder so that it won’t use human files such as the .db file?

Best, Woody

On Tue, Oct 31, 2023 at 6:54 AM ktruve @.***> wrote:

Hi again,

yes that is right. We get the sorted file for human. I see now that we get an error message at the end of the run with Danio_rerio: Reading... PSIsigma.db (ERROR) 153345 can't find wings in the database. Aborting... ===Filtering spent 0.0000 hours.=== Archiving... PSIsigma_r10_ir3.txt PSIsigma_r10_ir3.txt

What does that mean? Best,

Katarina

— Reply to this email directly, view it on GitHub https://github.com/wososa/PSI-Sigma/issues/53#issuecomment-1786975171,

or unsubscribe < https://github.com/notifications/unsubscribe-auth/ABWLHWKQHKVRRJAAXZE2PXTYCDKETAVCNFSM6AAAAAA6WTLMGWVHI2DSMVQWIX3LMV43OSLTON2WKQ3PNVWWK3TUHMYTOOBWHE3TKMJXGE>

. You are receiving this because you commented.Message ID: @.***>

— Reply to this email directly, view it on GitHub< https://github.com/wososa/PSI-Sigma/issues/53#issuecomment-1786981113>, or unsubscribe< https://github.com/notifications/unsubscribe-auth/A5JMZXXMRSCNKD5VLTPDL4LYCDKTHAVCNFSM6AAAAAA6WTLMGWVHI2DSMVQWIX3LMV43OSLTON2WKQ3PNVWWK3TUHMYTOOBWHE4DCMJRGM>.

You are receiving this because you were mentioned.Message ID: @.***>

— Reply to this email directly, view it on GitHub https://github.com/wososa/PSI-Sigma/issues/53#issuecomment-1787304151, or unsubscribe https://github.com/notifications/unsubscribe-auth/ABWLHWNHPDTKQYVXN5JIPATYCEBTXAVCNFSM6AAAAAA6WTLMGWVHI2DSMVQWIX3LMV43OSLTON2WKQ3PNVWWK3TUHMYTOOBXGMYDIMJVGE . You are receiving this because you commented.Message ID: @.***>

ktruve commented 1 year ago

Hi,

we reran and got the same results. We deleted all previous output files before running including .db, so that should not be the problem.

Any other suggestions what could go wrong?

Also would it be safe to use the results from PSIsigma_r10_ir3.txt.tar.gz, or what happens in the last step when the sorted file is created?

Best,

Katarina

On 1 Nov 2023, at 01:52, Woody Lin @.***> wrote:

Hi Katarina,

Could you delete the PSIsigma.db and run again? It is likely that the .db file wasn’t generated correctly.

Thanks, Woody

On Tue, Oct 31, 2023 at 10:14 AM ktruve @.***> wrote:

Hi,

We have Danio_rerio in a separate folder so that could not be the problem. Also deleted everything and reran, but same problem.

Any other suggestions?

Best,

Katarina

Statistics option = Student's t-test

number of p-value = 13378

Number of final p-value = 13121

Skipping p-value adjustment.

number of fdr(BH) = 13121

===PSI analysis spent 0.0119 hours.===

Filtering ΔPSI results...

Filtering mode = 0

Reading... /medstore/databases/research/Danio_rerio/Ensembl/Danio_rerio.GRCz11.110.chr.sorted.gtf.mapping.txt

Reading... PSIsigma.db

(ERROR) 153345 can't find wings in the database.

Aborting...

===Filtering spent 0.0000 hours.===

Archiving... PSIsigma_r10_ir3.txt

PSIsigma_r10_ir3.txt

On 31 Oct 2023, at 11:57, Woody Lin @.***> wrote:

Hi Katarina,

This means what you might be using the wrong PSIsigma.db file. Could you please try running PSI-Sigma for Danio_rerio in a separate folder so that it won’t use human files such as the .db file?

Best, Woody

On Tue, Oct 31, 2023 at 6:54 AM ktruve @.***> wrote:

Hi again,

yes that is right. We get the sorted file for human. I see now that we get an error message at the end of the run with Danio_rerio: Reading... PSIsigma.db (ERROR) 153345 can't find wings in the database. Aborting... ===Filtering spent 0.0000 hours.=== Archiving... PSIsigma_r10_ir3.txt PSIsigma_r10_ir3.txt

What does that mean? Best,

Katarina

— Reply to this email directly, view it on GitHub https://github.com/wososa/PSI-Sigma/issues/53#issuecomment-1786975171,

or unsubscribe < https://github.com/notifications/unsubscribe-auth/ABWLHWKQHKVRRJAAXZE2PXTYCDKETAVCNFSM6AAAAAA6WTLMGWVHI2DSMVQWIX3LMV43OSLTON2WKQ3PNVWWK3TUHMYTOOBWHE3TKMJXGE>

. You are receiving this because you commented.Message ID: @.***>

— Reply to this email directly, view it on GitHub< https://github.com/wososa/PSI-Sigma/issues/53#issuecomment-1786981113>, or unsubscribe< https://github.com/notifications/unsubscribe-auth/A5JMZXXMRSCNKD5VLTPDL4LYCDKTHAVCNFSM6AAAAAA6WTLMGWVHI2DSMVQWIX3LMV43OSLTON2WKQ3PNVWWK3TUHMYTOOBWHE4DCMJRGM>.

You are receiving this because you were mentioned.Message ID: @.***>

— Reply to this email directly, view it on GitHub https://github.com/wososa/PSI-Sigma/issues/53#issuecomment-1787304151, or unsubscribe https://github.com/notifications/unsubscribe-auth/ABWLHWNHPDTKQYVXN5JIPATYCEBTXAVCNFSM6AAAAAA6WTLMGWVHI2DSMVQWIX3LMV43OSLTON2WKQ3PNVWWK3TUHMYTOOBXGMYDIMJVGE . You are receiving this because you commented.Message ID: @.***>

— Reply to this email directly, view it on GitHubhttps://github.com/wososa/PSI-Sigma/issues/53#issuecomment-1788233126, or unsubscribehttps://github.com/notifications/unsubscribe-auth/A5JMZXTF4PP55GTF22X3WELYCGMLHAVCNFSM6AAAAAA6WTLMGWVHI2DSMVQWIX3LMV43OSLTON2WKQ3PNVWWK3TUHMYTOOBYGIZTGMJSGY. You are receiving this because you were mentioned.Message ID: @.***>

wososa commented 1 year ago

Hi @ktruve ,

Yes, it is possible to use PSIsigma_r10_ir3.txt.tar.gz as long as you filter results by yourself because the sorted.txt file is simply a filtered version of it. However, it is concerning to me that you can get sorted.txt for human only. It implies that there is something unexpected in the Danio_rerio.GRCz11.110.chr.sorted.gtf. I am not sure if this will compromise the results in the PSIsigma_r10_ir3.txt.tar.gz.

Does any of the results in the PSIsigma_r10_ir3.txt.tar.gz make sense to you? For example, if you pick an event with dPSI>20 and p-value < 0.01, do you see that event in the sashimi plot of IGV by using your alignment files from STAR?

Best, Woody

wososa commented 1 year ago

Hi @ktruve ,

Has your issue been solved?

Thanks, Woody

ktruve commented 1 year ago

Hi,

I’m not sure. There seem to be events at the positions checking with stringtie/IGV, but very hard to know it the calculations of p-values etc are correct, (?)

What does the error message mean?

Best,

Katarina

On 3 Nov 2023, at 13:56, Woody Lin @.***> wrote:

Hi @ktruvehttps://github.com/ktruve ,

Has your issue been solved?

Thanks, Woody

— Reply to this email directly, view it on GitHubhttps://github.com/wososa/PSI-Sigma/issues/53#issuecomment-1792393316, or unsubscribehttps://github.com/notifications/unsubscribe-auth/A5JMZXW2SVRMVBVIVOFJCKTYCTSXDAVCNFSM6AAAAAA6WTLMGWVHI2DSMVQWIX3LMV43OSLTON2WKQ3PNVWWK3TUHMYTOOJSGM4TGMZRGY. You are receiving this because you were mentioned.Message ID: @.***>

wososa commented 1 year ago

Hi @ktruve ,

The error message is saying that one of the records can't find matched coordinates in the PSIsigma.db file. This usually means that the PSisigma.db file wasn't generated correctly. It could be that PSI-Sigma failed to correctly parse the .gtf file you provided. Could you share a few lines of your .gtf file?

Best, Woody

wososa commented 1 year ago

Hi @ktruve ,

If the problem has been solved, I will close this thread.

Best, Woody