In recent months, when anyone in lab has been trying to run Diamond BLAST for functional annotations, permission errors or other errors have been an issue when trying to use the premade scripts to download the most recent nr database in FASTA format from NCBI and uses it to make a Diamond-formatted nr database on our Andromeda shared services.
Go to the sbatch_executables subdirectory in the Putnam Lab shared folder and run the scripts, make_diamond_nr_db.sh and make_diamond_nr_db.sh in the order below.
Looking into this, the permissions needed to be changed for these scripts as they were overall owned by Hollie, fixed that error. Then, I was still getting errors for downloading this step, but I found when I deleted the previous error and output file in the shared folder labeled output.4294967294.out and error.4294967294.out, the scripts ran.
This was because in these two scripts, the output and error files had %a ```#SBATCH -o output.%a.out
SBATCH -e error.%a.out ``` in it which becomes 4294967294 since this isn’t an array job. This should be %j to use the job ID, which will be unique each time.
Updated this in the scripts moving forward and they ran without issues.
In recent months, when anyone in lab has been trying to run Diamond BLAST for functional annotations, permission errors or other errors have been an issue when trying to use the premade scripts to download the most recent nr database in FASTA format from NCBI and uses it to make a Diamond-formatted nr database on our Andromeda shared services.
Go to the sbatch_executables subdirectory in the Putnam Lab shared folder and run the scripts,
make_diamond_nr_db.sh
andmake_diamond_nr_db.sh
in the order below.Looking into this, the permissions needed to be changed for these scripts as they were overall owned by Hollie, fixed that error. Then, I was still getting errors for downloading this step, but I found when I deleted the previous error and output file in the shared folder labeled output.4294967294.out and error.4294967294.out, the scripts ran.
This was because in these two scripts, the output and error files had %a ```#SBATCH -o output.%a.out
SBATCH -e error.%a.out ``` in it which becomes 4294967294 since this isn’t an array job. This should be %j to use the job ID, which will be unique each time.
Updated this in the scripts moving forward and they ran without issues.