bbuchfink
2.0.13 suffers with the eggNOG mapper
I try eggNOG mapper v2.1.6 with Diamond 2.0.13. The computer crashs down or restarts when eggNOG mapper calls the Diamond. I try to run Diamond according to eggNOG mapper, the computer suffers again.
However, I just finished a blastx alignment with Diamond 2.0.13 to Nr database. Then I change to use eggNOG mapper calls the Diamond 2.0.11 and it finished then. So there is must be something wrong in Diamond 2.0.13. My system is Linux Mint 20.0 with Xfct desktop.
What is your query file (type of sequences, sequence length, total size)? How much memory does your computer have? What is your command line for calling eggnog mapper or diamond?
My computer has 64Gb memory. The query file is a fasta file of protein sequences with 17.2Mb optained by TransDecoder from a transcriptome assembly. The commond line:
Diamond 2.0.13： emapper.py --cpu 12 --data_dir (to my database location) -m diamond -i XXXX.pep --out XXX (computer restrated or crash down) emapper.py --cpu 12 --data_dir (to my database location) -m diamond -i XXXX.pep --out XXX --sensmode fast (computer restrated or crash down) diamond blastp -d /XXX/XXX/eggnog_proteins.dmnd -q /XXX/XXX/XXXX.pep --threads 12 -o /XXX/XXX/emapper.hits --sensitive --iterate -e 0.001 --top 3 --outfmt 6 (computer restrated or crash down)
I copy the eggNOG database to a SSD card then: emapper.py --cpu 12 --data_dir (to my database location) -m diamond -i XXXX.pep --out XXX --sensmode fast (computer restrated or crash down)
diamond blastx to nr database with --sensitive (core dumped, The query file is a fasta of transcriptome with 76Mb, maybe the limitied memory is small) diamond blastx to nr database with --fast (completed, the query file is a fasta of transcriptome with 76Mb)
Diamond 2.0.11: emapper.py --cpu 12 --data_dir (to my database location) -m diamond -i XXXX.pep --out XXX (completed, default sensmode is sensitive.)
I have run a similarly sized dataset against the eggnog protein database using the command line that you provided, comparing versions 2.0.13 and 2.0.11. The runtime, CPU and memory usage was similar in both cases, so I can't reproduce a problem at the moment. Would it be possible to make your query file available to me so I can look into this further?