How to confirm SNPs results from a proteomics approach

[didipao]

Hi,

I used a mass spectrometry approach (de novo sequencing) and found single amino acid variations between 2 individuals at the protein level.

I want to test whether this variations are due to SNPs in the DNA or not.

I also need to be able to discriminate heterozygous vs. homozygous variations.

I have 2 samples and 20 "SNPs" to test.

Which method do you recommend?

I am no experience in genomics!

Thanks a lot!

[Richard Finney]

Do you know where this mutation is on a genome?

[didipao]

I know the location of the mutation in the protein, but I do not have the exact corresponding location in the genome.

[chadn737]

If you know with confidence which gene the peptides are coming from, just sequence that gene. Obtain DNA samples. Design primer pairs that will cover the region from which your peptide originated and then sequence the PCR product. You can then compare your two samples and see if there is a difference in the sequence.

[dphansti]

Sounds like an interesting project. Considering the recent paper highlighting RNA editing you might want to sequence the transcripts as well if you don't find the differences at the DNA level.

Also, I am guessing if you are asking the question you don't have access to a PCR machine etc. The scope of this task is fairly small so you should be able to find a collaborator to help.

[didipao]

@chadn737: Yes, i know with confidence which gene the peptides are coming from. Is the strategy you are proposing a Sanger sequencing. Do you know if I will be able to differentiate heterozygous from homozygous mutations?

[didipao]

@dphansti: It's very interesting, since de novo sequencing by mass spectrometry is very novel! Do yo have the reference for the RNA editing paper?

Actually, we want to sequence RNA as well. We want to find out at which level the mutations are arising.

I have access to PCR machine, but we do not have experience with genomics.

[chadn737]

Yes, I am proposing sanger sequencing. For the small number of snps you are wanting to deal with, it is the most sensible approach. You don't even need to resequence the entire gene, just the region from which your peptide originated.

You can distinguish heterozygous mutations in the trace files as long as you have good sequence. If you do find a snp in the DNA, then there are other methods you can use to further confirm this, particularly if its a heterozygous.

[didipao]

Thanks for your advice!

[gringer]

If you're just looking for confirmation using an alternative method, you don't even need to sequence:

http://en.wikipedia.org/wiki/SNP_genotyping#PCR-based_methods

You can design sequence-specific primers for the detected SNP variants, and pair them with reverse primers that are noticeably different distances from the SNP (i.e. 4 primers in total per SNP). You then run them on a gel, and count presence/absence of particular bands (heterozygous individuals will have both bands). With a sufficiently high resolution gel, you should be able to handle 20 different SNPs in one lane. Just remember your positive controls (you may be able to get away with both samples pooled together) and negative controls.

[chadn737]

If you're just looking for confirmation using an alternative method, you don't even need to sequence:

http://en.wikipedia.org/wiki/SNP_genotyping#PCR-based_methods

You can design sequence-specific primers for the detected SNP variants, and pair them with reverse primers that are noticeably different distances from the SNP (i.e. 4 primers in total per SNP). You then run them on a gel, and count presence/absence of particular bands (heterozygous individuals will have both bands). With a sufficiently high resolution gel, you should be able to handle 20 different SNPs in one lane. Just remember your positive controls (you may be able to get away with both samples pooled together) and negative controls.

That works assuming you know the sequence of the snp, however according to didipao these were identified using proteomics, so the underlying DNA sequence may not be known and still have to be sequenced.

[gringer]

That works assuming you know the sequence of the snp, however according to didipao these were identified using proteomics, so the underlying DNA sequence may not be known and still have to be sequenced.

This is a good point. I just assumed didipao had access to the gene sequence from this statement:

Yes, i know with confidence which gene the peptides are coming from.

[dphansti]

didipao. Here is the paper about RNA editing. Pretty interesting paper. They validated some of the editing events they found with mass spec.

Li M, Wang IX, Li Y, Bruzel A, Richards AL, Toung JM, Cheung VG. Widespread RNA and DNA Sequence Differences in the Human Transcriptome. Science. 2011 May19. [Epub ahead of print] PubMed PMID: 21596952.

[didipao]

didipao. Here is the paper about RNA editing. Pretty interesting paper. They validated some of the editing events they found with mass spec.

Li M, Wang IX, Li Y, Bruzel A, Richards AL, Toung JM, Cheung VG. Widespread RNA and DNA Sequence Differences in the Human Transcriptome. Science. 2011 May19. [Epub ahead of print] PubMed PMID: 21596952.

thanks, it looks very interesting!

[didipao]

thanks, it looks very interesting