bwa index problem when indexing files with large size

[544728460]

I have downloaded the nt data base from NCBI , for which the size is about 90G, I want to create a index file with commands as follows: bwa index -a bwtsw nt.fasta

Although it is running without any problem but it seems that the iterations can not come to an end after several days running.

I think it may be the problem of the big size for nt.fasta file. So is there any other methods for me to index big files like nt.fasta?

Thanks!

[MonaLiu421]

Hello,sorry to bother you. I have the same problem, have you solved it?

[544728460]

Sorry, i haven't solved it yet and i just chose to split the large file into several small files before indexing.

[MonaLiu421]

I‘m so glad to get your reply, I would like to ask is it reasonable to split the reference like this? Will it affect the result of mapping?

[544728460]

Yes, it's definitely reasonable because splitting large files is not only what i did to solve this problem but NCBI did the same thing with their nt database. (For details, see https://ftp.ncbi.nlm.nih.gov/blast/db/)

[MonaLiu421]

ok, thank you,you are so nice.

[lh3]

You may try a larger -b value

bwa index -b 100000000

It will be a little faster, but indexing nt will take days anyway.

[544728460]

Okay, thanks! I'll try later.

[bogv127515]

I also encountered the same problem. As you suggested, I increased the original -b parameter by a factor of 10. Unfortunately, fna.pac still doesn't grow after reaching the same size as before. The log file does not record the occurrence of the error. I think this may be caused by the memory constraints of my compute nodes, and simply increasing the -b parameter may not help much.