Identify Snps For Bacteria - Annotation

Question

[Akki]

Hello,

I have some 22 SNPs (from bacteria - using samtools pileup). I want to predict whether these occur in coding region, and if they do what type of substitutions are they (synonymous, non-synonymous etc). The format is:

Ref_chr01 1176674 G A 14 14 60 4 ..t^~A 4%%H

Any suggestions?

[Larry_Parnell]

A highly relevant question given the recent events in Germany. You use the word "predict" and that implies that there is no genome sequence for your organism. Do you want to predict SNP location/function or learn about SNP location/function given that there is a solved genome sequence?

Please clarify your question. Thanks!

Answer 1

[Michael]

Effect type prediction for prokaryotes is relatively simple in comparison to eukaryotes because there is no splicing. UCSC genome browser doesn't help you with prokaryote genomes, but you do not need very sophisticated tools anyway.

You need:

• The genome reference sequence for alignment (you have it)
• The genome annotation to get the coordinates of CDS (simply as a .gff file)

You can find the genome annotations for prokaryotes on the NCBI ftp server For the yersinia pestis KIM10 strain this is e.g.: ftp://ftp.ncbi.nih.gov/genbank/genomes/Bacteria/Yersinia_pestis_KIM_10_uid288/AE009952.gff

Make sure the strain is consistent throughout your analysis.

Then:

1. Find the CDS from the gff file for each SNP coordinate if any.
2. extract the reference sequence of all CDS that contain a predicted SNP
3. translate the reference sequence
4. insert the alternative allele at the predicted SNP position
5. translate again, if the AA sequence is different then consequence is non-synonymous, and synonymous otherwise

Hope this helps.

Answer 2

[Hx]

You can try using Annovar, I guess it's a helpful tool where it can annotate variations against UCSC or Refseq gene

I'm not sure whether there is genome sequence for your organism just like Larry mentioned.

You can try annotating your SNPs or indels if your genome sequence is present in the UCSC genome browser to identify whether SNPs cause protein coding changes and the amino acids that are affected. Annovar is open-source:

http://www.openbioinformatics.org/annovar/annovar_faq.html

Answer 3

[Swbarnes2]

Just 22 SNPs?

You could do that manually. Get your SNP in context, with maybe 100 bases on either side, and submit it to ncbi's blastx

http://blast.ncbi.nlm.nih.gov/Blast.cgi?PROGRAM=blastx&BLAST_PROGRAMS=blastx&PAGE_TYPE=BlastSearch&SHOW_DEFAULTS=on&LINK_LOC=blasthome

It'll translate your sequence in all 6 frames, and figure out which protein submitted to NCBI best fits. If you have found a SNP whose protein has been submitted, that hit will come up first, so use some caution. But in general, if your SNP causes a non-synonymous change, you'll see it by looking at the top alignment.

Answer 4

[Akki]

Thank you for your answer.

My genome is y. pestis (CO92). I have identified some SNPs and their locations - I want to cross-validate whether the SNPs at these positions are synonymous or non-synonymous. Also, I have a reference sequence of this species as a fasta file. I am relatively new to SNP calling- and still in learning phase. I apologize if these questions are trivial !!!

Thanks.

[Larry_Parnell]

Akki, in the future, please attach the above as a "comment" to the answer. Also, if it is a good answer, then give it a vote +1.

[Akki]

I apologize... I will keep it in mind :) !

Answer 5

[a.a.kechin]

You can use very simple and useful program Procannot: http://www.platform.genexplain.com/procannot/

Answer 6

[mkasamber]

hi all of you I want to annotate the SNPs of my genome P.aeruginosa (free software, easy and for windows).