search

prody / ProDy

A Python Package for Protein Dynamics Analysis
http://prody.csb.pitt.edu
Other
419 stars 154 forks source link

pca analysis error #1442

Closed TianmingQu closed 2 years ago

TianmingQu commented 2 years ago

@> Building covariance [ 99%] 1s

@> Building covariance [ 99%] 1s

@> Building covariance [ 99%] 1s

@> Building covariance [ 99%] 1s

@> Building covariance [ 99%] 1s

@> Building covariance [ 99%] 1s

@> Building covariance [ 99%] 1s

@> Building covariance [ 99%] 1s

@> Building covariance [ 99%] 1s

@> Building covariance [ 99%] 1s

@> Building covariance [ 99%] 1s

@> Building covariance [ 99%] 1s

@> Building covariance [ 99%] 1s

@> Building covariance [ 99%] 1s

@> Building covariance [ 99%] 1s

@> Building covariance [ 99%] 1s

@> Building covariance [ 99%] 1s

@> Building covariance [ 99%] 1s

@> Building covariance [ 99%] 1s

@> Building covariance [ 99%] 1s

@> Building covariance [ 99%] 1s

@> Building covariance [ 99%] 1s

@> Building covariance [ 99%] 1s

@> Building covariance [ 99%] 1s

@> Building covariance [ 99%] 1s

@> Building covariance [ 99%] 1s

@> Building covariance [ 99%] 1s

@> Building covariance [ 99%] 1s

@> Building covariance [ 99%] 1s

@> Building covariance [ 99%] 1s

@> Building covariance [ 99%] 1s

@> Building covariance [ 99%] 1s

@> Building covariance [ 99%] 1s

@> Building covariance [ 99%] 1s

@> Building covariance [ 99%] 1s

@> Building covariance [ 99%] 1s

@> Building covariance [ 99%] 1s

@> Building covariance [ 99%] 1s

@> Building covariance [ 99%] 1s

@> Building covariance [ 99%] 1s

@> Building covariance [ 99%] 1s

@> Building covariance [ 99%] 1s

@> Building covariance [ 99%] 1s

@> Building covariance [ 99%] 1s

@> Building covariance [ 99%] 1s

@> Building covariance [ 99%] 1s

@> Building covariance [ 99%] 1s

@> Building covariance [ 99%] 1s

@> Building covariance [ 99%] 1s

@> Building covariance [ 99%] 1s

@> Building covariance [ 99%] 1s

@> Building covariance [ 99%] 1s

@> Building covariance [ 99%] 1s

@> Building covariance [ 99%] 1s

@> Building covariance [ 99%] 1s

@> Building covariance [ 99%] 1s

@> Building covariance [ 99%] 1s

@> Building covariance [ 99%] 1s

@> Building covariance [ 99%] 1s

@> Building covariance [ 99%] 1s

@> Building covariance [ 99%] 1s

@> Building covariance [ 99%] 1s

@> Building covariance [ 99%] 1s

@> Building covariance [ 99%] 1s

@> Building covariance [ 99%] 1s

@> Building covariance [ 99%] 1s

@> Building covariance [ 99%] 1s

@> Building covariance [ 99%] 1s

@> Building covariance [ 99%] 1s

@> Building covariance [ 99%] 1s

@> Building covariance [ 99%] 1s

@> Building covariance [ 99%] 1s

@> Building covariance [ 99%] 1s

@> Building covariance [ 99%] 1s

@> Building covariance [ 99%] 1s

@> Building covariance [ 99%] 1s

@> Building covariance [100%] 0s while executing "exec $::NMWiz::pybin $::NMWiz::prody {*}$args" (procedure "::NMWiz::prodySubmitPCAjob" line 61) invoked from within "::NMWiz::prodySubmitPCAjob" (procedure "::NMWiz::prodySubmitJob" line 26) invoked from within "::NMWiz::prodySubmitJob" invoked from within ".nmwizprody.submitFrame.prodySubmit invoke" ("uplevel" body line 1) invoked from within "uplevel #0 [list $w invoke]" (procedure "tk::ButtonUp" line 23) invoked from within "tk::ButtonUp .nmwizprody.submitFrame.prodySubmit" (command bound to event)

Hi all, I am new to this plugin in vmd and I am trying to do pca analysis on prody, I saw some people having this issue before, could anyone give me some suggestions on how should I fix it? Thanks a lot!!

jamesmkrieger commented 2 years ago

Are you sure you’re using the latest ProDy (version 2.0.1)? I definitely have fixed this problem recently.

Best wishes James

TianmingQu commented 2 years ago

Are you sure you’re using the latest ProDy (version 2.0.1)? I definitely have fixed this problem recently.

Best wishes James

Hi James,

Thanks a lot for your so fast reply, I just checked my install information it is 2.0.1. So this really confused me, it would be better if you have any further idea on how to solve this problem.

Best, Tianming

TianmingQu commented 2 years ago

Are you sure you’re using the latest ProDy (version 2.0.1)? I definitely have fixed this problem recently.

Best wishes James

Hi, I just tried it with another protein, it seems that the ProDY PCA calculation can work for some protein and some does not work, could this be related with the size of protein and trajectory file?

Best, Tianming

jamesmkrieger commented 2 years ago

Shouldn't be related to something like that. Did you use the same VMD window for different proteins or could there be some difference in how it was configured?

jamesmkrieger commented 2 years ago

I'm thinking that maybe your VMD might sometimes see a different ProDy to the 2.0.1 version you checked

TianmingQu commented 2 years ago

Yes, I use the same vmd window. Interesting thing is that, when I use the terminal command “prody pca xxx.dcd” it is able to calculate the covariance matrix and generate the xxx.nmd file, which can be loaded into vmd using normal mode wizard, however with the same protein prody interface fail to generate the pca result in vmd.

Tianming

发件人: James Krieger @.> 答复: prody/ProDy @.> 日期: 2022年1月13日 星期四 上午11:18 收件人: prody/ProDy @.> 抄送: Tianming Qu @.>, Author @.***> 主题: Re: [prody/ProDy] pca analysis error (Issue #1442)

Shouldn't be related to something like that. Did you use the same VMD window for different proteins or could there be some difference in how it was configured?

— Reply to this email directly, view it on GitHubhttps://urldefense.com/v3/__https:/github.com/prody/ProDy/issues/1442*issuecomment-1012290146__;Iw!!Epnw_ITfSMW4!8mL4LM-Oqhgwqtg-mr7c-7WtDR1EPNk5QiukFnvrLn8XlmPVFyWzNmqGhQ63chBdGDI$, or unsubscribehttps://urldefense.com/v3/__https:/github.com/notifications/unsubscribe-auth/AQXNBSRVB3N732OXTZ4PBYTUV33NPANCNFSM5L2LCZEA__;!!Epnw_ITfSMW4!8mL4LM-Oqhgwqtg-mr7c-7WtDR1EPNk5QiukFnvrLn8XlmPVFyWzNmqGhQ630MkRp9Q$. Triage notifications on the go with GitHub Mobile for iOShttps://urldefense.com/v3/__https:/apps.apple.com/app/apple-store/id1477376905?ct=notification-email&mt=8&pt=524675__;!!Epnw_ITfSMW4!8mL4LM-Oqhgwqtg-mr7c-7WtDR1EPNk5QiukFnvrLn8XlmPVFyWzNmqGhQ639KlHDGg$ or Androidhttps://urldefense.com/v3/__https:/play.google.com/store/apps/details?id=com.github.android&referrer=utm_campaign*3Dnotification-email*26utm_medium*3Demail*26utm_source*3Dgithub__;JSUlJSU!!Epnw_ITfSMW4!8mL4LM-Oqhgwqtg-mr7c-7WtDR1EPNk5QiukFnvrLn8XlmPVFyWzNmqGhQ63Psmnxzk$. You are receiving this because you authored the thread.Message ID: @.***>

TianmingQu commented 2 years ago

You mean my vmd is using a different prody version?

Tianming

发件人: James Krieger @.> 答复: prody/ProDy @.> 日期: 2022年1月13日 星期四 上午11:19 收件人: prody/ProDy @.> 抄送: Tianming Qu @.>, Author @.***> 主题: Re: [prody/ProDy] pca analysis error (Issue #1442)

I'm thinking that maybe your VMD might sometimes see a different ProDy to the 2.0.1 version you checked

— Reply to this email directly, view it on GitHubhttps://urldefense.com/v3/__https:/github.com/prody/ProDy/issues/1442*issuecomment-1012290575__;Iw!!Epnw_ITfSMW4!-QWPNxPimFxwGtOR9D1QpjX5RhwRdkRKsnnX4DE1F8Tp1PkZh2D3H6CYBx3g2JSSe2Q$, or unsubscribehttps://urldefense.com/v3/__https:/github.com/notifications/unsubscribe-auth/AQXNBSRP2ZG5CYCYYNHBB6TUV33PJANCNFSM5L2LCZEA__;!!Epnw_ITfSMW4!-QWPNxPimFxwGtOR9D1QpjX5RhwRdkRKsnnX4DE1F8Tp1PkZh2D3H6CYBx3gz21uWRE$. Triage notifications on the go with GitHub Mobile for iOShttps://urldefense.com/v3/__https:/apps.apple.com/app/apple-store/id1477376905?ct=notification-email&mt=8&pt=524675__;!!Epnw_ITfSMW4!-QWPNxPimFxwGtOR9D1QpjX5RhwRdkRKsnnX4DE1F8Tp1PkZh2D3H6CYBx3g2xlQdhk$ or Androidhttps://urldefense.com/v3/__https:/play.google.com/store/apps/details?id=com.github.android&referrer=utm_campaign*3Dnotification-email*26utm_medium*3Demail*26utm_source*3Dgithub__;JSUlJSU!!Epnw_ITfSMW4!-QWPNxPimFxwGtOR9D1QpjX5RhwRdkRKsnnX4DE1F8Tp1PkZh2D3H6CYBx3gGONE8LQ$. You are receiving this because you authored the thread.Message ID: @.***>

jamesmkrieger commented 2 years ago

Yes, I use the same vmd window. Interesting thing is that, when I use the terminal command “prody pca xxx.dcd” it is able to calculate the covariance matrix and generate the xxx.nmd file, which can be loaded into vmd using normal mode wizard, however with the same protein prody interface fail to generate the pca result in vmd.

This is a way of getting around the problem. VMD doesn't like the logging text output generated by the prody pca command, which is also used by the interface. If you just run the command outside VMD then this logging text isn't seen by it, so that makes sense.

jamesmkrieger commented 2 years ago

You mean my vmd is using a different prody version?

Could be. Can you open the tk console from the extensions menu in vmd and type prody --version and see if this also says ProDy 2.0.1 please. Can you also type which prody so we can see where the prody it's using is.

Can you provide me information about how you checked that you had prody 2.0.1 too please

Hopefully we can work this out

TianmingQu commented 2 years ago

Alright I guess that’s one way I can use it. Another thing is that, if I only want to use the CA atom of residue number 100-200 on one protein to calculate covariance matrix, do you know how I should change the command in terminal?

I tried : (prody pca xxx.dcd –select “resid 100 to 200 and name CA”), but it seems that all the atoms were counted into the calculation again.

Tianming

发件人: James Krieger @.> 答复: prody/ProDy @.> 日期: 2022年1月13日 星期四 上午11:41 收件人: prody/ProDy @.> 抄送: Tianming Qu @.>, Author @.***> 主题: Re: [prody/ProDy] pca analysis error (Issue #1442)

Yes, I use the same vmd window. Interesting thing is that, when I use the terminal command “prody pca xxx.dcd” it is able to calculate the covariance matrix and generate the xxx.nmd file, which can be loaded into vmd using normal mode wizard, however with the same protein prody interface fail to generate the pca result in vmd.

This is a way of getting around the problem. VMD doesn't like the logging text output generated by the prody pca command, which is also used by the interface. If you just run the command outside VMD then this logging text isn't seen by it, so that makes sense.

— Reply to this email directly, view it on GitHubhttps://urldefense.com/v3/__https:/github.com/prody/ProDy/issues/1442*issuecomment-1012310293__;Iw!!Epnw_ITfSMW4!7_UMOsaWBCZ9IbjjDyxx0fJp6AHyMsYbJ563XgFnTAPkqzN1a1D4tkExDDat90gt2t8$, or unsubscribehttps://urldefense.com/v3/__https:/github.com/notifications/unsubscribe-auth/AQXNBSV6V7264UKB3I2M3FDUV36BVANCNFSM5L2LCZEA__;!!Epnw_ITfSMW4!7_UMOsaWBCZ9IbjjDyxx0fJp6AHyMsYbJ563XgFnTAPkqzN1a1D4tkExDDatndELyyw$. Triage notifications on the go with GitHub Mobile for iOShttps://urldefense.com/v3/__https:/apps.apple.com/app/apple-store/id1477376905?ct=notification-email&mt=8&pt=524675__;!!Epnw_ITfSMW4!7_UMOsaWBCZ9IbjjDyxx0fJp6AHyMsYbJ563XgFnTAPkqzN1a1D4tkExDDatkE597qE$ or Androidhttps://urldefense.com/v3/__https:/play.google.com/store/apps/details?id=com.github.android&referrer=utm_campaign*3Dnotification-email*26utm_medium*3Demail*26utm_source*3Dgithub__;JSUlJSU!!Epnw_ITfSMW4!7_UMOsaWBCZ9IbjjDyxx0fJp6AHyMsYbJ563XgFnTAPkqzN1a1D4tkExDDatFIXk26I$. You are receiving this because you authored the thread.Message ID: @.***>

jamesmkrieger commented 2 years ago

I think that should be basically right except you need to use resnum instead of resid in prody

TianmingQu commented 2 years ago

[文本 描述已自动生成]

Hi, I attached the screenshot with this email, it is 2.0.1.

Best, Tianming

发件人: James Krieger @.> 答复: prody/ProDy @.> 日期: 2022年1月13日 星期四 上午11:44 收件人: prody/ProDy @.> 抄送: Tianming Qu @.>, Author @.***> 主题: Re: [prody/ProDy] pca analysis error (Issue #1442)

You mean my vmd is using a different prody version?

Could be. Can you open the tk console from the extensions menu in vmd and type prody --version and see if this also says ProDy 2.0.1 please. Can you also type which prody so we can see where the prody it's using is.

Can you provide me information about how you checked that you had prody 2.0.1 too please

Hopefully we can work this out

— Reply to this email directly, view it on GitHubhttps://urldefense.com/v3/__https:/github.com/prody/ProDy/issues/1442*issuecomment-1012313358__;Iw!!Epnw_ITfSMW4!_0wFACDGrCQx-e6Gt4ZUt-xXWXGAF28jeK8z7ZvduWBd7_dCzctLL_Q4OG3UjYIXdvs$, or unsubscribehttps://urldefense.com/v3/__https:/github.com/notifications/unsubscribe-auth/AQXNBSVACXEPFV2IHKKPKDTUV36PLANCNFSM5L2LCZEA__;!!Epnw_ITfSMW4!_0wFACDGrCQx-e6Gt4ZUt-xXWXGAF28jeK8z7ZvduWBd7_dCzctLL_Q4OG3UvhILulo$. Triage notifications on the go with GitHub Mobile for iOShttps://urldefense.com/v3/__https:/apps.apple.com/app/apple-store/id1477376905?ct=notification-email&mt=8&pt=524675__;!!Epnw_ITfSMW4!_0wFACDGrCQx-e6Gt4ZUt-xXWXGAF28jeK8z7ZvduWBd7_dCzctLL_Q4OG3U-et5Fpc$ or Androidhttps://urldefense.com/v3/__https:/play.google.com/store/apps/details?id=com.github.android&referrer=utm_campaign*3Dnotification-email*26utm_medium*3Demail*26utm_source*3Dgithub__;JSUlJSU!!Epnw_ITfSMW4!_0wFACDGrCQx-e6Gt4ZUt-xXWXGAF28jeK8z7ZvduWBd7_dCzctLL_Q4OG3UlzKTGmg$. You are receiving this because you authored the thread.Message ID: @.***>

jamesmkrieger commented 2 years ago

Oh you probably need to provide a pdb or psf file too so it can find which atoms those are

jamesmkrieger commented 2 years ago

Hi, I attached the screenshot with this email, it is 2.0.1. Best, Tianming

The attachment doesn't end up on GitHub or in my emails. Can you send it to me directly at jamesmkrieger@gmail.com please?

TianmingQu commented 2 years ago

I tried : (prody pca xxx.pdb xxx.dcd --select "resid 100 to 200 and name CA") but it does not work

发件人: James Krieger @.> 答复: prody/ProDy @.> 日期: 2022年1月13日 星期四 上午11:47 收件人: prody/ProDy @.> 抄送: Tianming Qu @.>, Author @.***> 主题: Re: [prody/ProDy] pca analysis error (Issue #1442)

Oh you probably need to provide a pdb or psf file too so it can find which atoms those are

— Reply to this email directly, view it on GitHubhttps://urldefense.com/v3/__https:/github.com/prody/ProDy/issues/1442*issuecomment-1012315560__;Iw!!Epnw_ITfSMW4!_Z4OLCGPiajbfsd3IxSnEYbV5b0L7H8vDTdbuJuI3100lFNoNG_SE9bhRpGIi7WOHl8$, or unsubscribehttps://urldefense.com/v3/__https:/github.com/notifications/unsubscribe-auth/AQXNBSQYIG4BB3DHDGSSCEDUV36YZANCNFSM5L2LCZEA__;!!Epnw_ITfSMW4!_Z4OLCGPiajbfsd3IxSnEYbV5b0L7H8vDTdbuJuI3100lFNoNG_SE9bhRpGIAw5_Zhs$. Triage notifications on the go with GitHub Mobile for iOShttps://urldefense.com/v3/__https:/apps.apple.com/app/apple-store/id1477376905?ct=notification-email&mt=8&pt=524675__;!!Epnw_ITfSMW4!_Z4OLCGPiajbfsd3IxSnEYbV5b0L7H8vDTdbuJuI3100lFNoNG_SE9bhRpGIouTU7Cg$ or Androidhttps://urldefense.com/v3/__https:/play.google.com/store/apps/details?id=com.github.android&referrer=utm_campaign*3Dnotification-email*26utm_medium*3Demail*26utm_source*3Dgithub__;JSUlJSU!!Epnw_ITfSMW4!_Z4OLCGPiajbfsd3IxSnEYbV5b0L7H8vDTdbuJuI3100lFNoNG_SE9bhRpGIVDZtF6I$. You are receiving this because you authored the thread.Message ID: @.***>

jamesmkrieger commented 2 years ago

If you type prody pca --help then you get the following information:

usage: prody pca [-h] [--quiet] [--examples] [-n INT] [-s SEL] [-a] [-o PATH] [-e] [-r] [-u] [-q] [-v] [-z] [-t STR]
                 [-j] [-p STR] [-f STR] [-d STR] [-x STR] [-A] [-R] [-Q] [-J STR] [-F STR] [-D INT] [-W FLOAT]
                 [-H FLOAT] [--psf PSF | --pdb PDB] [--aligned]
                 dcd

positional arguments:
  dcd                   file in DCD or PDB format

optional arguments:
  -h, --help            show this help message and exit
  --quiet               suppress info messages to stderr
  --examples            show usage examples and exit
  --psf PSF             PSF filename
  --pdb PDB             PDB filename
  --aligned             trajectory is already aligned

parameters:
  -n INT, --number-of-modes INT
                        number of non-zero eigenvectors (modes) to calculate (default: 10)
  -s SEL, --select SEL  atom selection (default: "protein and name CA or nucleic and name P C4' C2")

I think you should try the following: prody pca xxx.dcd --pdb yyyy.pdb --select "name CA and resnum 100 to 200"

jamesmkrieger commented 2 years ago

I think you need to change resid to resnum

On Thu, Jan 13, 2022 at 4:49 PM TianmingQu @.***> wrote:

I tried : (prody pca xxx.pdb xxx.dcd --select "resid 100 to 200 and name CA") but it does not work

发件人: James Krieger @.> 答复: prody/ProDy @.> 日期: 2022年1月13日 星期四 上午11:47 收件人: prody/ProDy @.> 抄送: Tianming Qu @.>, Author @.***> 主题: Re: [prody/ProDy] pca analysis error (Issue #1442)

Oh you probably need to provide a pdb or psf file too so it can find which atoms those are

— Reply to this email directly, view it on GitHub< https://urldefense.com/v3/__https:/github.com/prody/ProDy/issues/1442*issuecomment-1012315560__;Iw!!Epnw_ITfSMW4!_Z4OLCGPiajbfsd3IxSnEYbV5b0L7H8vDTdbuJuI3100lFNoNG_SE9bhRpGIi7WOHl8$>, or unsubscribe< https://urldefense.com/v3/__https:/github.com/notifications/unsubscribe-auth/AQXNBSQYIG4BB3DHDGSSCEDUV36YZANCNFSM5L2LCZEA__;!!Epnw_ITfSMW4!_Z4OLCGPiajbfsd3IxSnEYbV5b0L7H8vDTdbuJuI3100lFNoNG_SE9bhRpGIAw5_Zhs$>.

Triage notifications on the go with GitHub Mobile for iOS< https://urldefense.com/v3/__https:/apps.apple.com/app/apple-store/id1477376905?ct=notification-email&mt=8&pt=524675__;!!Epnw_ITfSMW4!_Z4OLCGPiajbfsd3IxSnEYbV5b0L7H8vDTdbuJuI3100lFNoNG_SE9bhRpGIouTU7Cg$> or Android< https://urldefense.com/v3/__https:/play.google.com/store/apps/details?id=com.github.android&referrer=utm_campaign*3Dnotification-email*26utm_medium*3Demail*26utm_source*3Dgithub__;JSUlJSU!!Epnw_ITfSMW4!_Z4OLCGPiajbfsd3IxSnEYbV5b0L7H8vDTdbuJuI3100lFNoNG_SE9bhRpGIVDZtF6I$>.

You are receiving this because you authored the thread.Message ID: @.***>

— Reply to this email directly, view it on GitHub https://github.com/prody/ProDy/issues/1442#issuecomment-1012317725, or unsubscribe https://github.com/notifications/unsubscribe-auth/ADFFDGTCJD7BCTTAA7D2NW3UV37CBANCNFSM5L2LCZEA . Triage notifications on the go with GitHub Mobile for iOS https://apps.apple.com/app/apple-store/id1477376905?ct=notification-email&mt=8&pt=524675 or Android https://play.google.com/store/apps/details?id=com.github.android&referrer=utm_campaign%3Dnotification-email%26utm_medium%3Demail%26utm_source%3Dgithub.

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

jamesmkrieger commented 2 years ago

Yes, I use the same vmd window

Actually, I think it could make sense that if you make a small enough selection the logging just doesn't happen. There same was reported in #746

jamesmkrieger commented 2 years ago

Actually, looking back at my fix, I don't think it's general enough. It suppresses messages about superposition but not about building the covariance matrix. I'll try and make another fix

jamesmkrieger commented 2 years ago

I also did just find that there is a difference in what the prody pca app and interface do depending on whether the dcd is larger than 1000 frames or not so that could explain why it only sometimes happens.

jamesmkrieger commented 2 years ago

This should now be fixed. Please update ProDy from GitHub to take advantage of it. 

pip uninstall prody
pip install git+https://github.com/prody/ProDy.git
jamesmkrieger commented 2 years ago

This should finally be fixed in #1552. You're absolutely right that it wasn't properly fixed in 2.0.1