Genotyping 20 genes, 30 snps

[Cinsculpta]

What would be the most cost-efficient way to genotyping ~20 genes at 30 SNPs from 5ml human saliva sample?
1. Real-time PCR with fluorescent reporter (Allele-Specific PCR)
2. Mass Spec
3. Customized chip
4. NGS (assume hitchhiking with a busy HiSeq 2500)

Then, what if sequencing the whole exon of those 20 genes?

Really appreciate any input

[gringer]

The low number of SNPs and genes rules out (4) for typing.

How many samples? If it's only a handful, it's unlikely that a custom chip (3) will be cost effective.

I would expect that the cheapest would be the mass spec genotyping.

Depending on the exon lengths, a tiled PCR and standard Sanger sequencing might be the cheapest for full exon sequencing. If not, then a small portion of a HiSeq lane, say 5%, should give you more than enough coverage.

[Cinsculpta]

Thank you David for your input. I'd say 100-500 samples...Would Real-time PCR with fluorescent reporter be out of the game? Is it possible to come up with a ballpark of how much these methods would cost each? Thanks again.

The low number of SNPs and genes rules out (4) for typing.

How many samples? If it's only a handful, it's unlikely that a custom chip (3) will be cost effective.

I would expect that the cheapest would be the mass spec genotyping.

Depending on the exon lengths, a tiled PCR and standard Sanger sequencing might be the cheapest for full exon sequencing. If not, then a small portion of a HiSeq lane, say 5%, should give you more than enough coverage.

[gringer]

A whole exome beadChip will be about $100 per sample (plus sample prep, additional reagents, and profit margins):

http://www.illumina.com/products/infinium_humanexome_beadchip_kit.html

A custom-designed chip will probably be cheaper; but you'd need to talk to the array suppliers about that:

http://www.illumina.com/products/infinium_iselect_custom_genotyping_beadchips.html

Sequenom at IDDRC (just happened to be the first result I could find with an actual cost) is about $2000 for 30 SNPs and 384 samples (~$5 per sample):

http://core.iddrc.org/molecular-genetics/?page_id=46

LGC Genomics seems to be a similar price.

You could also have a look here for more ideas:

http://seqanswers.com/forums/showthread.php?t=32357

I'm less sure about the rtPCR cost.

[nucacidhunter]

Amplicon sequencing on MiSeq is another option. Three options:
1- TruSeq Custom amplicon
2- Custom Accel-amplicon (Swift Biosciences)
3- In house design following Illumina two-step amplicon sequencing

[gringer]

Yes, of course. I just thought MiSeq was out because Cinsculpta mentioned a HiSeq 2500.

[Cinsculpta]

Thanks again, David! (need to find out the availability of these fancy machines)

A whole exome beadChip will be about $100 per sample (plus sample prep, additional reagents, and profit margins):

http://www.illumina.com/products/infinium_humanexome_beadchip_kit.html

A custom-designed chip will probably be cheaper; but you'd need to talk to the array suppliers about that:

http://www.illumina.com/products/infinium_iselect_custom_genotyping_beadchips.html

Sequenom at IDDRC (just happened to be the first result I could find with an actual cost) is about $2000 for 30 SNPs and 384 samples (~$5 per sample):

http://core.iddrc.org/molecular-genetics/?page_id=46

LGC Genomics seems to be a similar price.

You could also have a look here for more ideas:

http://seqanswers.com/forums/showthread.php?t=32357

I'm less sure about the rtPCR cost.

[Cinsculpta]

Sorry for my delayed reply...Thanks a lot for the info. What about the cost per sample (assuming MiSeq available)?

Amplicon sequencing on MiSeq is another option. Three options:
1- TruSeq Custom amplicon
2- Custom Accel-amplicon (Swift Biosciences)
3- In house design following Illumina two-step amplicon sequencing

[TonyBrooks]

My guess would be 1. Especially if you had a 384 well fluorescent plate reader. It works out at less than $0.50 per data point ($15 per sample).

[nucacidhunter]

If you want to genotype 30 SNP across 500 samples that will be 1500 amplicon. MiSeq v3 reagents output would at least 20M reads. To compensate for possible low diversity and to increase quality of reads one can lower cluster density in which case the output will be around 15M. So the average coverage will be 10000/amplicon. The major cost will be preparing amplicon libraries. If you choose commercial suppliers job will be easy (couple of days at bench) and success is guaranteed. They can design probes and synthesise them and supply all required reagents as a kit. You might get a quote from commercial sources in your region. Input is 50-250 ng for custom amplicon kits.

[ithersam]

PCR can also be used for genotyping when trying to compare fragments of different lengths; for instance, short tandem repeat (STR) typing. STR typing is normally performed by labeling one primer of each primer pair

[yuhj]

Why not use Sequennom platform

[gringer]

I believe that would be point 2 of the initial post. I mentioned a price for that (about $5 per sample, according to IDDRC).