Hi there! Which dependencies are you referring to specifically? There is no conda enviroment for RaFAH. Also, which version of RaFAH are you running and with which command?
If you would like to refer to this comment somewhere else in this project, copy and paste the following link:
Hello, I think the comment refers to the Docker image as I also get the same error when running the Docker image "--predict --merged_cds_file_name". Do you have any suggestions? Thanks, Julia
If you would like to refer to this comment somewhere else in this project, copy and paste the following link:
We can also see an issue (as in the original comment) using the test data - TS3_Toy_Genomes.fasta (downloaded from here https://sourceforge.net/projects/rafah/files/Data/Toy_Set.tgz/download). If I download the data to my "docker" folder and run the command as described in the readme I get this:
Ok. And there is no "RaFAH_1_Host_Predictions.tsv" file generated, correct? It seems R is dying after loading the models into memory. This could happen if there is not enough RAM available. RaFAH needs at least 80 Mb of RAM to run properly, so it will most likely not work on a personal computer. Can you try running it on your server and let me know if you still get the same error?
If you would like to refer to this comment somewhere else in this project, copy and paste the following link:
Thanks for the message. Tracking the memory usage of the Docker container appears to go up to 8Gigs and then is closer to 3 Gigs memory use when it has the file not found error.
If you would like to refer to this comment somewhere else in this project, copy and paste the following link:
My bad, I meant 80GB (considering a dataset with thousands of viral sequences and using the predict mode). It only uses this much for a very short time, but enough to get killed if there is not enough much memory available.
If you would like to refer to this comment somewhere else in this project, copy and paste the following link:
Ok thanks. I ran it on AWS and it was fine. I think it would be helpful for users to know min requirements (or to improve the error handling). I cannot make a PR here on Sourceforge or else I would have submitted some propositions. Thanks for your help, Julia
If you would like to refer to this comment somewhere else in this project, copy and paste the following link:
Thanks Julia, I will add min requirements to the readme file. Also, I just created a RaFAH repo on github: https://github.com/felipehcoutinho/RaFAH Its about time I center everything there anyway. Feel free to submit PRs through there. Best, F
If you would like to refer to this comment somewhere else in this project, copy and paste the following link:
Hi there! Which dependencies are you referring to specifically? There is no conda enviroment for RaFAH. Also, which version of RaFAH are you running and with which command?
Hello, I think the comment refers to the Docker image as I also get the same error when running the Docker image "--predict --merged_cds_file_name". Do you have any suggestions? Thanks, Julia
Hi Julia, can you please specify the full command line you are using?
Thanks for the response!
docker run --mount type=bind,source=/Users/juliagustavsen/Documents/viral_analysis/results,target=/results/ fhcoutinho/rafah perl /usr/bin/RaFAH.pl --predict --merged_cds_file_name /results/Total_CP.fasta --output_dir /results
Is "/results/Total_CP.fasta" empty by any chance? If there is something in it, are they protein sequences in fasta format?
yes they are protein sequences in fasta format.
We can also see an issue (as in the original comment) using the test data - TS3_Toy_Genomes.fasta (downloaded from here https://sourceforge.net/projects/rafah/files/Data/Toy_Set.tgz/download). If I download the data to my "docker" folder and run the command as described in the readme I get this:
docker run --mount type=bind,source=/Users/juliagustavsen/Documents/thesis_projects/Kuwait_gp20_analysis/docker,target=/results/ fhcoutinho/rafah perl /usr/bin/RaFAH.pl --predict --genomes_dir /results/ --extension .fasta --file_prefix RaFAH_1
Ok. And there is no "RaFAH_1_Host_Predictions.tsv" file generated, correct? It seems R is dying after loading the models into memory. This could happen if there is not enough RAM available. RaFAH needs at least 80 Mb of RAM to run properly, so it will most likely not work on a personal computer. Can you try running it on your server and let me know if you still get the same error?
Thanks for the message. Tracking the memory usage of the Docker container appears to go up to 8Gigs and then is closer to 3 Gigs memory use when it has the file not found error.
My bad, I meant 80GB (considering a dataset with thousands of viral sequences and using the predict mode). It only uses this much for a very short time, but enough to get killed if there is not enough much memory available.
Ok thanks. I ran it on AWS and it was fine. I think it would be helpful for users to know min requirements (or to improve the error handling). I cannot make a PR here on Sourceforge or else I would have submitted some propositions. Thanks for your help, Julia
Thanks Julia, I will add min requirements to the readme file. Also, I just created a RaFAH repo on github: https://github.com/felipehcoutinho/RaFAH Its about time I center everything there anyway. Feel free to submit PRs through there. Best, F