SEQanswers

SEQanswers (http://seqanswers.com/forums/index.php)
-   Bioinformatics (http://seqanswers.com/forums/forumdisplay.php?f=18)
-   -   Taxonomy classifier (http://seqanswers.com/forums/showthread.php?t=10655)

Chuckytah 04-10-2011 02:46 PM

Taxonomy classifier
 
Hello,

I have found this website: http://rdp.cme.msu.edu/classifier/classifier.jsp but it's for bacteria and archeobacteria...

don't know about one for all the kingdoms?

I have a contigs.fasta and i want to know how many sequences i have from wich taxonomic groups...

bckirkup 04-10-2011 05:00 PM

Greengenes at Lawrence Livermore will help you with that, in concert with ARB.

Chuckytah 04-11-2011 06:11 AM

Quote:

Originally Posted by bckirkup (Post 39178)
Greengenes at Lawrence Livermore will help you with that, in concert with ARB.


there is some website? thank you so much.

Greengenes + ARB?

boetsie 04-11-2011 06:45 AM

I think he means this website;

http://greengenes.lbl.gov/cgi-bin/nph-index.cgi

Chuckytah 04-11-2011 07:07 AM

Quote:

Originally Posted by boetsie (Post 39234)

thank you. Someone advised me to use MEGAN... what you think?

Palecomic 04-11-2011 07:16 AM

I'd also check out Silva http://www.arb-silva.de/ it's from the guys who made Arb. The plus points are that the alignment is better than greengenes (which uses an old alignment from the Silva group) or RDP (which doesn't bother aligning the variable regions of 16S), the negative points are that aligning more than 300 sequences requires paying for their SINA aligner. You can get round this by aligning using Mothur from Pat Schloss http://www.mothur.org/, doing the alignment there and then merging it with the current Silva database.

You mention that you have a contigs file, which suggests that you're looking at metagenomics data, so perhaps you want to try something like MG RAST http://metagenomics.anl.gov/ or PhylOTU http://www.ploscompbiol.org/article/...l.pcbi.1001061 which will get you phylogenetic information from a metagenomic rather than 16S dataset?

Palecomic 04-11-2011 07:20 AM

Sorry, just spotted your update. MEGAN would do the job, but check out that PhylOTU paper for a brief discussion of its limitations.

Chuckytah 04-11-2011 08:45 AM

Quote:

Originally Posted by Palecomic (Post 39239)
I'd also check out Silva http://www.arb-silva.de/ it's from the guys who made Arb. The plus points are that the alignment is better than greengenes (which uses an old alignment from the Silva group) or RDP (which doesn't bother aligning the variable regions of 16S), the negative points are that aligning more than 300 sequences requires paying for their SINA aligner. You can get round this by aligning using Mothur from Pat Schloss http://www.mothur.org/, doing the alignment there and then merging it with the current Silva database.

You mention that you have a contigs file, which suggests that you're looking at metagenomics data, so perhaps you want to try something like MG RAST http://metagenomics.anl.gov/ or PhylOTU http://www.ploscompbiol.org/article/...l.pcbi.1001061 which will get you phylogenetic information from a metagenomic rather than 16S dataset?



Silva http://www.arb-silva.de/ seems nice but i have more than 300 sequences :(

Chuckytah 04-11-2011 09:13 AM

Quote:

Originally Posted by Palecomic (Post 39240)
Sorry, just spotted your update. MEGAN would do the job, but check out that PhylOTU paper for a brief discussion of its limitations.

Megan asks me for a .RMA file... but i only have .txt and . fasta and .BlastClust and excel files... there is some conversion tool?

Chuckytah 04-14-2011 07:30 AM

Quote:

Originally Posted by Chuckytah (Post 39247)
Megan asks me for a .RMA file... but i only have .txt and . fasta and .BlastClust and excel files... there is some conversion tool?

Still don't know what to do :(

how can i have a .RMA file?

themerlin 04-15-2011 05:34 AM

This is from the manual:

Quote:

To prepare a dataset for use with MEGAN, one must first compare the given reads against a database of reference sequences, for example by performing a BLASTX search against the NCBI-NR database. The file of reads and the resulting BLAST file can then be directly imported into MEGAN and the program will automatically calculate a taxonomic classification of the reads and also, if desired, a functional classification, using either the SEED or KEGG classification, or both.
You import the blasta data to process it, but you can load processed megan results in .rma format.

Chuckytah 04-15-2011 08:09 AM

Quote:

Originally Posted by themerlin (Post 39585)
This is from the manual:


You import the blasta data to process it, but you can load processed megan results in .rma format.

I will try it later :) ty so much

Chuckytah 04-24-2011 06:58 PM

Quote:

Originally Posted by themerlin (Post 39585)
This is from the manual:


You import the blasta data to process it, but you can load processed megan results in .rma format.

I don't have any blasta file... only fasta

JackieBadger 04-25-2011 03:03 AM

Quote:

Originally Posted by Chuckytah (Post 39174)
Hello,

I have found this website: http://rdp.cme.msu.edu/classifier/classifier.jsp but it's for bacteria and archeobacteria...

don't know about one for all the kingdoms?

I have a contigs.fasta and i want to know how many sequences i have from wich taxonomic groups...

The Galaxy web portal has a MegaBlast tool, which will also build trees of your sequences highlighted by taxa. Watch the video tutorial on "Processing and analysis of 454 data in metagenomic studies"

http://main.g2.bx.psu.edu/

Chuckytah 07-10-2011 08:41 AM

there are some tutorials for Megan???

Chuckytah 10-29-2011 12:37 PM

Quote:

Originally Posted by JackieBadger (Post 40222)
The Galaxy web portal has a MegaBlast tool, which will also build trees of your sequences highlighted by taxa. Watch the video tutorial on "Processing and analysis of 454 data in metagenomic studies"

http://main.g2.bx.psu.edu/

Sorry, dunno why only saw your comment today. I will now watch the tutorial.

pbluescript 11-02-2011 12:55 PM

Thanks for the comments everyone! I just downloaded MEGAN and got it working in just a couple minutes. I should always just check seqanswers first before trying to find something on my own...


All times are GMT -8. The time now is 10:36 AM.

Powered by vBulletin® Version 3.8.9
Copyright ©2000 - 2019, vBulletin Solutions, Inc.