how to assign taxonomy for functional gene

[Lynn]

Dear all,

I would like to assign taxonomy the functional gene (cbbM). For 16S, I use RDP. For that gene, I don't know how to do. Could you please help me? If I need to use BLAST, I don't know how to install BLAST in my qiime1. I am using miniconda to use qiime. To blast 1 by 1 in blast website, my rep.seq are more than 10000.

Thanks.

best regards

lynn

[Lynn]

I tried to install blast in my miniconda environment using this command

sudo apt-get install ncbi-blast+

It appears error like this

Reading package lists... Done

Building dependency tree

Reading state information... Done

The following NEW packages will be installed:

  ncbi-blast+

0 upgraded, 1 newly installed, 0 to remove and 202 not upgraded.

128 not fully installed or removed.

Need to get 0 B/8,973 kB of archives.

After this operation, 38.6 MB of additional disk space will be used.

Setting up install-info (6.1.0.dfsg.1-5) ...

/usr/sbin/update-info-dir: 3: /etc/environment: Syntax error: Unterminated quoted string

dpkg: error processing package install-info (--configure):

 subprocess installed post-installation script returned error exit status 2

Errors were encountered while processing:

 install-info

E: Sub-process /usr/bin/dpkg returned an error code (1)

 Please help me how to solve this problem.

thanks

[TonyWalters]

Hello,

I will preface this with the general disclaimer that QIIME1 isn't updated anymore, and support is quite limited. I'd recommending moving to QIIME2 unless there is a specific reason for using QIIME1 (e.g. replicating an analysis).

For your questions:

The links to the old blast executables aren't valid anymore. You can download them here though:

https://ftp.ncbi.nlm.nih.gov/blast/executables/legacy.NOTSUPPORTED/2.2.22/

and get the blast-2.2.22-ia32-linux.tar.gz or blast-2.2.22-universal-macosx.tar.gz depending upon whether you have linux or a Mac.

The blast approach to doing taxonomic assignment will be the most lenient (compared to RDP or uclust), but you've still got to have two reference files:

1. a fasta file with your reference database sequences

2. a taxonomic mapping file with (tab delimited) with sequence labels matching the fasta labels in #1 followed by a tab, followed by the taxonomy strings.

I hope this helps,

Tony

[Lynn]

Dear Mr. Tony Walters,

Thank you so much for your reply. I installed blast in qiime environment and move directory to the folder where my sequence files stored. Now, It's ok. I got OTU table. 

Thanks again for your help.

with best regards

lynn

[Lynn]

Dear Mr. Tony,

I am sorry to disturb you again. But, I would like to ask you how to reduce the similarity level when we use blast to assign taxonomy.

I got more than 500 sequences as no blast hit. I already increased e value to 0.01. The results are same. 

I am looking forward to your reply.

thanks.

lynn

[TonyWalters]

Hello Lynn,

Unfortunately, the only setting for blast is the e value. Apart from moving the e value to a higher level, you might try blasting some of the reads that failed to hit manually on the NCBI nr site, and see if those reads hit the target gene of interest or if they're non-target amplification products.

-Tony

[Lynn]

Dear Mr. Tony,

Thank you so much for your answer. I will blast in NCBI page. 

with best regards

lynn