BLAST comparison between two sets of sequences

Peppe · 10-03-2014, 08:22 AM

Hello everyone,
I have a set of sequences generated with Trinity and I want to compare them with S. cerevisiae. I have downloaded the S. cerevisiae protein set from the SGD. Since S. cerevisiae is a reference organism, I need to name my transcripts using S. cerevisiae nomenclature.
For the comparison I used BLASTx. Although the BLASTx itself is quite easy to do using the Windows version of BLAST 2.2.29+, I run into the problem that I cannot get S. cerevisiae gene names.

As an example, one protein is indicated as
>YAL001C TFC3 SGDID:S000000001, Chr I from 151006-147594,151166-151097, Genome Release 64-1-1, reverse complement, Verified ORF, "Largest of six subunits of the RNA polymerase III transcription initiation factor complex (TFIIIC); part of the TauB domain of TFIIIC that binds DNA at the BoxB promoter sites of tRNA and similar genes; cooperates with Tfc6p in DNA binding"

I run BLASTx using several flags (eg -outfmt 6 or -outfmt 7 -qseqid -sseqid -stitle), but the only thing that I get is the systematic name (i.e. YAL001C) but not the gene name and the description (i.e. TFC3 etc.).

Have anybody run into the same problem? Do you guys have any advice? Doing manually the blast one by one of ~1000 sequences is going to drive me nuts!
Thanks a lot

amitm · 10-03-2014, 09:13 AM

hi Peppe,
YAL001C is a locus tag. A dirty solution would be to get the Blast result and then do an ID conversion from Locus tag to gene symbol.
Gene anno. info is available for all Fungi from NCBI ftp -
ftp://ftp.ncbi.nlm.nih.gov/gene/DATA...i.gene_info.gz

The file has symbols for all locus tags.
The first col of the file is the taxon ID. For Yeast that is 559292

Peppe · 10-03-2014, 09:26 AM

Thanks amitm.
In the way you said should I do the conversion manually? Or is there an informatic way?
Best.

amitm · 10-03-2014, 09:50 AM

hi,
Once you have a list of Locus Tag IDs, the gene symbol can be retrieved from the above file by using vlookup formula in excel.
The easiest way would have been to use an online ID conversion tool. But two that I know (DAVID and CNIO) either didn't recognize the locus tag ID or didn't support Yeast.

I would have gone with excel. Least hand-on time

Peppe · 10-03-2014, 01:56 PM

Thanks amitm,
I did and it worked fine. Now I have gene names for all my transcripts. By change do you know how to have also Saccharomyces gene descriptions? This would be ideal.

I think in BLAST someone can use several flags to select whatever he wants, but for some reason it does not work for my data sets.
Thanks.

amitm · 10-03-2014, 04:43 PM

hi Peppe,
Did you use the file from NCBI ftp? It has the gene description also, as in the full name, if thats what you meant. If you are looking for possible functions then it would be gene ontology. If you go up to the parent directory in the ftp, i.e. - ftp://ftp.ncbi.nlm.nih.gov/gene/DATA/

The gene2go.gz file has GO description for every gene. You could filter this file based on the GeneID (2nd col).

Alternatively, since you have the gene symbols, you could use online GO classification tools like DAVID (http://david.abcc.ncifcrf.gov/tools.jsp)

Similar Threads
Thread	Thread Starter	Forum	Replies	Last Post
BLAST comparison between two sets of sequences	Peppe	Introductions	1	10-03-2014 08:25 AM
comparing large sets of sequences	gsgs	Bioinformatics	11	01-10-2013 05:28 AM
Comparison of two sets of miRNAs	Giorgio C	Bioinformatics	6	03-30-2012 07:47 AM
MAnorm: a robust model for quantitative comparison of ChIP-Seq data sets	z3mao	Epigenetics	0	03-20-2012 05:30 PM

10-03-2014, 08:22 AM	#1
Peppe Member Location: USA Join Date: Nov 2012 Posts: 11	BLAST comparison between two sets of sequences Hello everyone, I have a set of sequences generated with Trinity and I want to compare them with S. cerevisiae. I have downloaded the S. cerevisiae protein set from the SGD. Since S. cerevisiae is a reference organism, I need to name my transcripts using S. cerevisiae nomenclature. For the comparison I used BLASTx. Although the BLASTx itself is quite easy to do using the Windows version of BLAST 2.2.29+, I run into the problem that I cannot get S. cerevisiae gene names. As an example, one protein is indicated as >YAL001C TFC3 SGDID:S000000001, Chr I from 151006-147594,151166-151097, Genome Release 64-1-1, reverse complement, Verified ORF, "Largest of six subunits of the RNA polymerase III transcription initiation factor complex (TFIIIC); part of the TauB domain of TFIIIC that binds DNA at the BoxB promoter sites of tRNA and similar genes; cooperates with Tfc6p in DNA binding" I run BLASTx using several flags (eg -outfmt 6 or -outfmt 7 -qseqid -sseqid -stitle), but the only thing that I get is the systematic name (i.e. YAL001C) but not the gene name and the description (i.e. TFC3 etc.). Have anybody run into the same problem? Do you guys have any advice? Doing manually the blast one by one of ~1000 sequences is going to drive me nuts! Thanks a lot

10-03-2014, 09:13 AM	#2
amitm Member Location: Manchester, UK Join Date: Feb 2011 Posts: 52	hi Peppe, YAL001C is a locus tag. A dirty solution would be to get the Blast result and then do an ID conversion from Locus tag to gene symbol. Gene anno. info is available for all Fungi from NCBI ftp - ftp://ftp.ncbi.nlm.nih.gov/gene/DATA...i.gene_info.gz The file has symbols for all locus tags. The first col of the file is the taxon ID. For Yeast that is 559292

10-03-2014, 09:26 AM	#3
Peppe Member Location: USA Join Date: Nov 2012 Posts: 11	Thanks amitm. In the way you said should I do the conversion manually? Or is there an informatic way? Best.

10-03-2014, 09:50 AM	#4
amitm Member Location: Manchester, UK Join Date: Feb 2011 Posts: 52	hi, Once you have a list of Locus Tag IDs, the gene symbol can be retrieved from the above file by using vlookup formula in excel. The easiest way would have been to use an online ID conversion tool. But two that I know (DAVID and CNIO) either didn't recognize the locus tag ID or didn't support Yeast. I would have gone with excel. Least hand-on time

10-03-2014, 01:56 PM	#5
Peppe Member Location: USA Join Date: Nov 2012 Posts: 11	Thanks amitm, I did and it worked fine. Now I have gene names for all my transcripts. By change do you know how to have also Saccharomyces gene descriptions? This would be ideal. I think in BLAST someone can use several flags to select whatever he wants, but for some reason it does not work for my data sets. Thanks.

10-03-2014, 04:43 PM	#6
amitm Member Location: Manchester, UK Join Date: Feb 2011 Posts: 52	hi Peppe, Did you use the file from NCBI ftp? It has the gene description also, as in the full name, if thats what you meant. If you are looking for possible functions then it would be gene ontology. If you go up to the parent directory in the ftp, i.e. - ftp://ftp.ncbi.nlm.nih.gov/gene/DATA/ The gene2go.gz file has GO description for every gene. You could filter this file based on the GeneID (2nd col). Alternatively, since you have the gene symbols, you could use online GO classification tools like DAVID (http://david.abcc.ncifcrf.gov/tools.jsp)