Problems creating a blast database

Question

[lucasmacielbiotec]

Hello,

I'm trying to annotate a transcriptome using de Blast2go. For this I downloaded the nr ncbi database to make the blast. My problem is that the file has 72GB and when I created the database using the command line "makeblastdb" the file is divided into 57 parts (as you can see at image nr.00, nr.01, nr.02....nr56 [https://ibb.co/bxBhBQ] ) but I have got to have just one part.

Anyone knows how to solve this problem?

Thanks !

[biofalconch]

I have never used Blast2go, but in normal blast this is normal, when you specify just the name of the database, I presume in this case it is nr. For example, if I were to run blastn it would be blastn -query foo.fasta -db nr

[lucasmacielbiotec]

The problem is that the program is not by command line and the program allows to chose just one file.

[Jake Warner]

I have also never used blast2go (too slow). Did you try using the index file nr.pal ?

Answer 1

[biomonte]

The free version of BLAST2GO can only be run through a graphical interface and is going to take forever on such a big dataset. Nowadays there are better alternatives to functionally annotate a transcriptome. I suggest you trying TRINOTATE which is very easy to run and it will give you comments from a bunch of different databases (e.g. GO, KEGG, Pfam, etc.).

[lucasmacielbiotec]

I'm trying the Trinotate but I found an error that I couldn't solve when I have to load the results into a Trinotate SQLite Database.

perl Trinotate Trinotate.sqlite init --gene_trans_map Trinity.fasta.gene_trans_map --transcript_fasta prasiola.fasta --transdecoder_pep longest_orfs.pep

CMD: /home/lucas/Downloads/Trinotate-3.0.2/admin/util/trinotateSeqLoader/TrinotateSeqLoader.pl --sqlite Trinotate.sqlite --gene_trans_map Trinity.fasta.gene_trans_map --transcript_fasta prasiola.fasta --transdecoder_pep longest_orfs.pep --bulk_load Can't exec "/home/lucas/Downloads/Trinotate-3.0.2/admin/util/trinotateSeqLoader/TrinotateSeqLoader.pl": Arquivo ou diretório não encontrado at Trinotate line 124. Error, cmd: /home/lucas/Downloads/Trinotate-3.0.2/admin/util/trinotateSeqLoader/TrinotateSeqLoader.pl --sqlite Trinotate.sqlite --gene_trans_map Trinity.fasta.gene_trans_map --transcript_fasta prasiola.fasta --transdecoder_pep longest_orfs.pep --bulk_load died with ret -1 at Trinotate line 126.

Have you ever seen this? Thank you for your attention.

[biofalconch]

Well, it is going to be hard since the output has some parts in portuguese (i think), but it says it couldn't find the file at "/home/lucas/Downloads/Trinotate-3.0.2/admin/util/trinotateSeqLoader/TrinotateSeqLoader.pl", you can check where it is and then modify the perl script :)

[biomonte]

Olá Lucas,

I would recommend you to please make sure you have everything available in your working directory. If you still have problems, you could try to automatically annotate your transcripts by using autoTrinotate.pl instead. I wrote a tutorial about this some time ago. You can find it here: https://github.com/biomendi/PRACTICAL-GUIDE-TO-BUILD-DE-NOVO-TRANSCRIPTOME-ASSEMBLIES-FOR-NON-MODEL-ORGANISMS/wiki/8.-Functional-annotation-using-TRINOTATE

Please let me know if everything is clear to you. Thank you!

Answer 2

[Michael]

I think you can use Blast2Go and give it a blast file generated by local Blast. However, to run local Blast please refer to the Blast manual and download blast databases only by using update_blastdb.pl --decompress nr

I have got to have just one part.

No, you need all the files!