njaupan / ecc_finder

a tool to detect eccDNA using Illumina and ONT sequencing
GNU General Public License v3.0
14 stars 5 forks source link

Installation error #17

Open hAmbreen opened 1 year ago

hAmbreen commented 1 year ago

Hi njaupan, I am trying to install ecc_finder using the provided yml file but it is throwing a lot of conflicts with incompatible packages and ultimately, the installation fails. I am running the code on Ubuntu server. I have tried the installation on a number of different servers but the problem remains true for all. Can you please help me identifying the underlying issue and how it could be surpassed. Also, is there is any other way to install the program other than using the yml file? I could not find one around.

Many thanks, Heena

njaupan commented 1 year ago

Hi Heena, The conflict came from conda version which yml file built in old 3,4. I will upload the new installation after holiday and keep you updated.

Best.

On Wed, Jan 4, 2023 at 8:51 AM hAmbreen @.***> wrote:

Hi njaupan, I am trying to install ecc_finder using the provided yml file but it is throwing a lot of conflicts with incompatible packages and ultimately, the installation fails. I am running the code on Ubuntu server. I have tried the installation on a number of different servers but the problem remains true for all. Can you please help me identifying the underlying issue and how it could be surpassed. Also, is there is any other way to install the program other than using the yml file? I could not find one around.

Many thanks, Heena

— Reply to this email directly, view it on GitHub https://github.com/njaupan/ecc_finder/issues/17, or unsubscribe https://github.com/notifications/unsubscribe-auth/AB5SXNZDJXVKQHH4DF25METWQV56TANCNFSM6AAAAAATQ2GDBE . You are receiving this because you are subscribed to this thread.Message ID: @.***>

hAmbreen commented 1 year ago

Thank you njaupan. Looking forward to the updated installation file.

Many Thanks, Heena

hAmbreen commented 1 year ago

Hi njaupan, Any update on the new installation file?

Thanks, Heena

lauramiozzi commented 1 year ago

Hi, I had the same problem; any update on the installation file? Thanks

Laura

silvia84 commented 1 year ago

Hello, same problem here. Any update on the new installation file?

Thank you, Silvia

julia-luz commented 10 months ago

Hello!

I have managed to install a partially working environment on linux - only for the map-ont function.

I have not been able to get it working for the asm-ontfunction or any or even tried the short read functions yet, and would really appreciate if anyone who has them working would be able to share their environment and any changes in the original code with us here, since Dr. PanPan has been inactive on github for a long time. A colleague of mine, who worked with her, has been trying to reach her for a while and will continue to do so, so fingers crossed.

In the mean time, if you would like to try my partial solution, download the attached file, change the extension from .txtto .ymlor .yamland then use it to follow the environment creation instructions as usual.

workingenv.txt

If anyone else who has this working as intended for other functions would kindly share their working environment, please use the conda env export --from-history function and share the file with us!

TingheZhang commented 6 months ago

I think this can be installed by using mamba instead of conda from yaml

julia-luz commented 6 months ago

I think this can be installed by using mamba instead of conda from yaml

Yep, it can also technically be done with conda + libmamba solver

kaufmanelij commented 6 months ago

I was able to install it using mamba after switching the tbb version in the .yaml file to the following: - tbb=2020.2=h4bd325d_4

It seem to run with the map-ont though even in the samples where ecc are detected the output FASTA are blank for some reason and it throws the following error: [E::stk_subseq] failed to open the input file/stream

julia-luz commented 6 months ago

@kaufmanelij

Hi! I ran into this problem a lot early on - I can't access my "proper" logs right now to be 100% sure, but if I remember correctly, your .fasta are empty because the .csv used to create the .fasta files are empty. This can happen because the input/output paths are incorrect or empty (easy to happen when trying to run this in a loop, or with bad corrupted files) or you ran out of RAM earlier on and the pipeline and have empty bam files, which in turn create empty bed files, which in turn create empty csv files, and finally empty fasta files... Usually the culprit is samtools, which at some point in this pipeline exits with error code 0, the rest just keeps on going with the bad input files until you get to seqtk. Genrich and bedtools also get troublesome in this pipeline sometimes.

Also, are you trying to run this with the overwrite flag? As written, there's a step in the code that ignores the overwrite tag and ignores new files, I'll try to post my fix for that later.

julia-luz commented 4 months ago

Since November, conda's default solver should be libmamba solver, so it should now be possible to install the pipeline.

yx-xu commented 4 months ago

Hi njaupan, I had the same problem, too. I am trying to install ecc_finder using the provided yml file with command conda env create -f ecc_finder.yaml, but found many conflicts with incompatible packages. I have tried the installation on python 2.7 or python3.4, but the problem still exsits. Could you please help me with this problem? Thank you very much!

yongxin

julia-luz commented 4 months ago

@yx-xu - Hello! Have you tried one of the solutions provided by others in this discussion?

Which version of conda are you using? The newest versions of conda should use libmamba as their default solver. Try updating your conda installation and trying again. If you cannot update conda, you can set libmamba as your default conda solver (https://conda.github.io/conda-libmamba-solver/user-guide/), do as @TingheZhang suggested and use Mamba instead, or use the minimal environment I shared above.

Let me know how it goes!

yx-xu commented 4 months ago

@julia-luz Hi julia! I have tried the installation on conda 23.9.1 or conda 4.14.0, they both found many conflicts with incompatible packages. I also installed conda-libmamba-solver and run conda env create -f ecc_finder.yaml --solver=libmamba but it results conflicts, too. Sorry for my limit analysis ability for bioinformatics data. Thanks for your help!
image

yongxin

yx-xu commented 4 months ago

@julia-luz Hello~ I changed the yaml file as kaufmanelij suggested - tbb=2020.2=h4bd325d_. And I got the result using example data. But the ecc.sr.csv and ecc.sr.fasta file were empty. I don't know if I ran it successfully? And which file should I focus on? Thank you very much!

image these are the resulte after run python ecc_finder.py map-sr, the ecc.sr.csv and ecc.sr.fasta file were empty. image but the file in directory peak_files/ have several line. image these are the resulte after run python ecc_finder.py asm-sr.

julia-luz commented 3 months ago

@julia-luz Hello~ I changed the yaml file as kaufmanelij suggested - tbb=2020.2=h4bd325d_. And I got the result using example data. But the ecc.sr.csv and ecc.sr.fasta file were empty. I don't know if I ran it successfully? And which file should I focus on? Thank you very much!

image these are the resulte after run python ecc_finder.py map-sr, the ecc.sr.csv and ecc.sr.fasta file were empty. image but the file in directory peak_files/ have several line. image these are the resulte after run python ecc_finder.py asm-sr.

Hey there! I have only played around with the long read options, so I'm not sure if I can help with the short read errors as I'm not as familiar with the sr code.

From my experience, you may have ran out of RAM and samtools crashed without giving you an error message. The resulting .bam file will be wrong, making the .bed files wrong/empty. The .csv files are generated from the .bed files, and the fasta is generated from the .csv, so they'll be empty.

Try running it again and log your output to a .txt file so you can go through each step and watch for any error message or warning, specially from samtools and/or Genrich.

Also, open a new terminal window while running the script and use htop to watch your RAM usage - you might see it rising to the max very quickly.

yx-xu commented 3 months ago

Hi Julia-luz! I have successfully ran the software and got results that seems to be right. Thanks for your help! Best wishes!

julia-luz commented 3 months ago

Hi Julia-luz! I have successfully ran the software and got results that seems to be right. Thanks for your help! Best wishes!

Great, I'm glad to know that! Would you be so kind as to share how you fixed your problem for others in need? Maybe open a new issue and reply to yourself with your solution, that would be fantastic.