-
Notifications
You must be signed in to change notification settings - Fork 1
New issue
Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.
By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.
Already on GitHub? Sign in to your account
returned non-zero exit status 255 for 5_epa_outgroup_rooting.py and IndexError: list index out of range for 6_root_digger_rooting.py #2
Comments
Hi Vinita, could you take a look in Pierre |
Hi @Pbdas I am attaching this Thank you very much Vinita |
@Pbdas Maybe we could just add This thread check is not that important in this context anyway |
Yes I agree, only data with ~30k sites will make it through the filters anyway. I just pushed the change to the master branch, so @vinitamehlawat you should be able to update by running (in the folder of the repository) As for the rootdigger stage, I'm not sure. @computations any idea? |
An even better solution would be upgrading to |
Hi @Pbdas Here I am working with my subset data which is consist of 2059 sequences After This time I encounter the following error
I also checked the I am attaching Thank you very much! |
Hi Vinita, the log file from RAxML-NG looks good now! The error seems to be with gappa this time. same procedure as last time: Pierre |
Hi @Pbdas After |
Hi @vinitamehlawat , just letting you know I think I figured out the current issue, and I'm working on a fix |
Ok, so I'm 99% sure the issue was that the call to Also, it could be that the next issue will be related to the visualization using R, meaning that it may be necessary to install some packages. Note however that this visualization is not strictly necessary and you can repeat it later by simply calling the script directly, like so:
(fill in the correct paths) the necessary R packages are:
|
Hi @Pbdas After Traceback (most recent call last): This time in Thanks |
Hi Vinita, Also please run this command in your terminal and paste the result here: Pierre |
Hi @Pbdas
This is free software; you are free to change and redistribute it.
|
Hi @vinitamehlawat, sorry it took a while, but could finally reproduce the issue on my end! Heres is what you do:
Now you can try stage 5 again. |
Hi @Pbdas Thank you , I made changes as per your suggestions. This time this shows error for Error in library(ggplot2) : there is no package called ‘ggplot2’ If this error is regarding to the only But could you please look at the
Traceback (most recent call last): Thank you very much for your time and effor Vinita |
Dear Vinita, It looks like I introduced a bug that we haven't detected. It should be fixed now. Best, |
I am little confuse, so after git pull I should run the 5th step again or
the whole analysis from step 1
…On Thu, 25 Nov 2021 at 5:52 PM, BenoitMorel ***@***.***> wrote:
Dear Vinita,
It looks like I introduced a bug that we haven't detected. It should be
fixed now.
Please try to run 'git pull' and to start the analysis again.
Let us know if that fixes the issue
Best,
Benoit
—
You are receiving this because you were mentioned.
Reply to this email directly, view it on GitHub
<#2 (comment)>,
or unsubscribe
<https://github.com/notifications/unsubscribe-auth/ALJQTD3VTHSLIAZ6E3IQ54LUNYTBVANCNFSM5II5W5TA>
.
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>.
|
Sorry, step 5 should be enough.
Le jeu. 25 nov. 2021 à 13:29, vinitamehlawat ***@***.***> a
écrit :
… I am little confuse, so after git pull I should run the 5th step again or
the whole analysis from step 1
On Thu, 25 Nov 2021 at 5:52 PM, BenoitMorel ***@***.***>
wrote:
> Dear Vinita,
>
> It looks like I introduced a bug that we haven't detected. It should be
> fixed now.
> Please try to run 'git pull' and to start the analysis again.
> Let us know if that fixes the issue
>
> Best,
> Benoit
>
> —
> You are receiving this because you were mentioned.
> Reply to this email directly, view it on GitHub
> <
#2 (comment)
>,
> or unsubscribe
> <
https://github.com/notifications/unsubscribe-auth/ALJQTD3VTHSLIAZ6E3IQ54LUNYTBVANCNFSM5II5W5TA
>
> .
> 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.
Reply to this email directly, view it on GitHub
<#2 (comment)>,
or unsubscribe
<https://github.com/notifications/unsubscribe-auth/ADJJ3UNQG6MAYBFMIIFZTXTUNYT3RANCNFSM5II5W5TA>
.
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>.
|
I replied too fast. My fix only fixes step 6. I don't think it depends on step 5. So you should rerun step 6 :-) |
Hi @BenoitMorel After
|
HI @Pbdas After installing all
hmmbuild :: profile HMM construction from multiple sequence alignments input alignment file: /home/vinita/covid19_cme_analysis/work_dir/2021-11-19_00/fmsan/data/covid_edited.fasta idx name nseq alen mlen W eff_nseq re/pos description 1 covid_edited 1807 27987 27920 29776 1.70 0.619 CPU time: 15.89u 0.32s 00:00:16.21 Elapsed: 00:00:16.20 Attaching package: ‘dplyr’ The following objects are masked from ‘package:stats’:
The following objects are masked from ‘package:base’:
Warning message: Simple named list: Auto named with Using lambdas Thank you |
Hi Vinita, the part that fails now is just a copy of the result files from As for the rootdigger error, in the Cheers, |
Hi @Pbdas I checked Thanks |
Hi Vinita, that makes debugging a lot harder. One thing I can think of right now is that MPI may not be installed on your machine, could it be? you can check by running |
Hi @Pbdas (base) Report bugs to http://www.open-mpi.org/community/help/ |
it is unfortunate that there is no log file, but there are things to try regardless. If you see a file called |
Ah, I found it (or at least what I think is going on). I pushed a fix to github for rootdigger. You will need to pull the new version and rebuild the program, and then you should be able to run the script successfully. |
Here I am again little confused, Could you please calrify after |
Ah, sorry, I should be more clear. in the top level directory, there should be a directory |
I did the same as you have mentioned above and it successfully updated the software but after re-run of 6th script I am getting same error ./pipeline/6_root_digger_rooting.py work_dir/2021-11-19_00/fmsan/ |
Alright, I think I managed to find the problem. There was a change in interface for rootdigger that didn't get updated in this pipeline. I have pushed a change to the pipeline, it should be good to just |
I did the running 8 iterations |
And there are still no logs at this time in
|
Yes, I still don't have |
Hi @Pbdas Thank you very much! Now I am able to ran 5th script without any error and got my outputs for this script. Again thank you for your time and efforts. Vinita |
It looks like the checkpoints might be corrupted. Remove any files in the |
I removed |
Hi @Pbdas I apologise for saying that the 5th script worked, but I got outputs for my three datasets fmsan, fmsao, and smsan, but (base) ./pipeline/5_epa_outgroup_rooting.py work_dir/2021-11-19_00/smsao/ Sorry for bothering you yet again |
@vinitamehlawat what happens when you remove |
@computations terminal now look like this (base) |
ok, and now try removing the checkpoint file and see what happens with that same command? |
@computations it is still running but not sure how much time it will take to compelet (base) /home/vinita/covid19_cme_analysis/software/root_digger/bin/rd --tree /home/vinita/covid19_cme_analysis/work_dir/2021-11-19_00/fmsan/runs/root_digger_runs/0.in.tree --msa /home/vinita/covid19_cme_analysis/work_dir/2021-11-19_00/fmsan/data/covid_edited.fasta --model GTR+FO+G4 --exhaustive --prefix /home/vinita/covid19_cme_analysis/work_dir/2021-11-19_00/fmsan/runs/root_digger_runs/0 --threads 48 |
Thanks for being patient. There is an estimated runtime there, and I find it to be approximately accurate. One thing to note, this is one of the 8 trees that would have been run. I am going to push a change to the script that fixes this so that you can just run the script. But, this will take a while, you have a very large tree. |
Hi Vinita, there should be a file called |
Hi @Pbdas Please find attached Thank you |
Hi Vinita, since you mentioned in #4 that for now you're only interested in getting a tree out, I think we can shelve this error for now. Placement is only relevant here if you want to try to see if it could find a better outgroup/root placement of the tree. Let me know if in the future you want this kind of analysis, then I'll have another look! Pierre |
Hi @Pbdas Thank you very much, But could you please suggest which scripts exactly I need to run to get thinned tree for my data and also as you mentioned Placement is enough so is its Vinita |
Hi Vinita, I will help you with the thinning. I am updating the wiki page to explain how it should be run, but I need some time to read the code and remember how to use it properly. I will come back to you as soon as possible Benoit |
Hi Vinita, I'm not sure I understand your question about placement, do you want to use it after all? ´scripts/placement.py´ contains functions that are use by pipeline stages. ´pipeline/wuhan_placement.py´ is a separate placement based stage that tries to place the original Wuhan SARS COV2 genome into the tree. If you're just interested in building a tree, you don't need placement (or rootdigger for that amtter) at all. Pierre |
Hi @Pbdas So thing is that before using your pipeline I was trying to construct a phylogenetic tree with two different softwares like RaXML and IQ-TREE but I didn't get good branch support in both trees, Then I read your paper which I found extremly helpful and followed your pipeline because you clealry mentioned that how difficult it is to do phylogeney for SARS-data with the low number of mutations in sequences. So My question is, in your pipeline which scripts are useful for my data (like Hope I am able to deliver my question |
Hi Vinita, I think the stages should be 0-3 to get the basic trees. Then stage 7 produces statistics about the trees and from that a set of "plausible" trees (these trees should be in a file called ´credible_ml_trees.newick´). From these, two consensus trees are built ( Note that a consensus tree (usually) has unresolved nodes, i.e. multifurcations. So you may need to resolve those depending on how you want to further use the tree. However usually these would be resolved randomly, so that doesn't give you any better information. The idea of using the set of plausible trees is that it lets you see the broader picture, which is why what we did here is to do any further steps once for each tree in that set, then look at the set of results and interpret them as a whole. I hope that makes sense! |
Hi @vinitamehlawat, I have updated the wiki page. I have made several minor changes, and added a section here: https://github.com/BenoitMorel/covid19_cme_analysis/wiki#description-of-the-steps If you want to run tree thinning, please first update your repo (there were a few bugs which are now fixed) and read the new wiki section. Let me know if that's unclear. Here is another important remark: we do not generate bootstrap trees in the current pipeline, and thus you won't get any support value (which is what interests you most :-)). covid19_cme_analysis/scripts/common.py Line 378 in 30245cf
After this, you need to restart from step 2, pargenes. But most importantly: now that I know what you want to achieve (and I am sorry to say this after all the painful debugging that you had to go through), I am not sure that this pipeline will help you more than just running raxml/iqtree. To get the support values, all we do is to run raxml-ng (all the other steps had some sense for our study, but not for your purpose). Best, |
Hi @BenoitMorel Thank you so much for well explained clarification, I will definately try [https://github.com/BenoitMorel/covid19_cme_analysis/blob/30245cf877552d0151ba73290042cd2d8e0eb7e4/scripts/common.py#L378] and tree thining . I apologise as well, @Pbdas @computations , for all of your efforts in debugging different scripts. Vinita |
Hi @idaios
I prepare my dataset from scratch having high quality 10 sars-cov2 genome with 2 outgruop so total sequences in my data are 12 for which I again ran all script, this time I am stuck at 5th script
When first time ran 5th script the ERROR was
:Then I ran 7th script first which is
7_iqtree_tests.py
and again ran 5th script which is giving following error./pipeline/5_epa_outgroup_rooting.py work_dir/2021-11-17_05/smsao
Further I tried 6th script
./pipeline/6_root_digger_rooting.py work_dir/2021-11-17_05/fmsan/
But somehow 8_tree_thinning.py, 9_mptp_on_all_trees.py, compare_llhs, and extract_thinned_dataset.py worked on 4 dataset which are FMSAO, SMSAO, FMSAN & SMSAN
For
wuhan_placement.py
also get some erro./pipeline/wuhan_placement.py work_dir/2021-11-17_05/smsao/
KIndly help me to understand these issue wether they are interlinked with my data or something which is not present in my data thats why root_digger_lwr.csv is empty in
smsao/results/rootdigger_rooting
It would be very great if you could look at these errors and suggest me how I should solve these.
Thank you very much
Vinita
The text was updated successfully, but these errors were encountered: