BLAST version in fragmentation tools for ab initio

[NooraAz]

There is a version incompatibility between the nr library and the BLAST version that causes the following error when running make_fragments.pl:

[fastacmd] ERROR: ERROR: Cannot initialize readdb for nr database

That's because the default nr library is upgraded to V5, but fastacmd can only read V4.
There are two options here:

1. Downgrading nr library
2. Upgrading BLAST

The new BLAST is faster, but it needs more work and rewriting of code. Downgrading nr is an easier fix.

I want to know your opinion, before working on it, should we downgrade nr or upgrade BLAST? And of course, I would appreciate any help in fixing this.

[dgront]

This question can be translated into: "how many people use the original nn_make script?" From my perspective, users typically used to get fragments from Robetta server, or from the newer version of frag_picker. If this is the case, I'd argue to downgrade the nr DB. Please also notice, that make_fragments.pl script calls secondary structure predictors (PsiPred, Juffo, etc) which also run blast. You'd have to ensure this still work after the upgrade. Best, Dominik Gront sob., 13 lip 2024 o 21:57 Noora Azadvari ***@***.***> napisał(a):

…

There is a version incompatibility between the *nr* library and the BLAST version that causes the following error when running make_fragments.pl: [fastacmd] ERROR: ERROR: Cannot initialize readdb for nr database That's because the default nr library is upgraded to V5 <https://www.biostars.org/p/436501/>, but fastacmd can only read V4. There are two options here: 1. Downgrading nr library 2. Upgrading BLAST The new BLAST is faster, but it needs more work and rewriting of code. Downgrading nr is an easier fix. I want to know your opinion, before working on it, should we downgrade nr or upgrade BLAST? And of course, I would appreciate any help in fixing this. — Reply to this email directly, view it on GitHub <#92>, or unsubscribe <https://github.com/notifications/unsubscribe-auth/AAC3MI5S3YADAT2MHVMYC53ZMGBBJAVCNFSM6AAAAABK2QZLROVHI2DSMVQWIX3LMV43ASLTON2WKOZSGQYDOMJUGA4TQNQ> . You are receiving this because you are subscribed to this thread.Message ID: ***@***.***>