Where can I get my whole genome sequenced and put in easy data format?

[steviek2000]

Hi,

I'm looking to get my whole genome sequenced. I researched the cost and it seems it can be done for 10k USD maximum.

 

After it is sequenced, is there software available that will easily show me all known genes that are in my dna ?

 

Maybe the software would compare it against a known database.

 

Is such a software system available that is free or not expensive for one license ?

 

I'm not a scientist so looking for layman type of software. My main concern is that I pay for it to get done but then it is not in any usable form or there software needed to analyze it and compare is to technical to use.

 

Thank You.

Edited July 23, 2013 by steviek2000

[ydoaPs]

From what I hear, 23andme has some good packages.

[Arete]

OK, so I'm going to guess what you mean by "whole genome sequenced for $10,000" is that you can get a DNA sample extracted, library prepped and run on a single lane of Illumina for around $10,000.

 

The data you will get from such an undertaking would be 150 - 220 million short read sequences either 50, 75 or 100 base pairs in length. The term "whole genome sequencing" is a bit of a misnomer, as they "should" - dependent on your library prep, be randomly sampled from the genome and as such you'll have higher than average coverage from some parts of the genome, and none from others.

 

De novo assembly is not possible from Illumina data alone, so the method you'd need to use would be a referenced alignment to an assembled genome, followed by SNP calling. By my back of the envelope, sketchy calculations, you'd be looking at around 3-8x average coverage from one lane of sequencing over the human genome, which isn't really enough to confidently SNP call - you'd want to get more like 20x. So your results would be fairly low quality in terms of confidence.

 

Second, while the alignment and SNP calling software is largely free (I like GATK, SAMtools and VCFtools) you need a fair bit of computing power to run the analyses efficiently, and some relatively basic programming skills in pipelining and big data. I currently use a high performance cluster which runs 48 processors and 512GB per node, and I'd get that type of alignment done in an hour or two.

 

A much, much better option if you simply want to know your allelic state at a bunch of important coding loci is a microarray approach. Affymetrix, Illumina, Fluidigm etc all have relatively cost effective, off the shelf SNP microarrays for humans which research scientists use all the time. As YodaPs points out, 23andme have a commercial service to run a variety of these chips for a lot less than 10 grand.

[steviek2000]

23andme seems to be the first place to go. I have ordered from them. Hopefully they will provide me with the raw microarrays.

 

Then if someone discovers are particular snp linked to a disease, inorder for me to test if it is in my dna, I first I need to have the loci microarray of where it would be and then need the software search for it.

 

For example I would like to be able to test to see if I have a certain SNP that a researcher has linked to a disease. ( given that 23andme doesn't already test for it ).

[CharonY]

Note that in area of biomarkers for many SNPs only association studies to diseases and conditions exist. I.e. there is no known mechanistic link between an allele and a given phenotype. Thus, many of the predictions are false positives, and I believe I read somewhere that 23andme heavily overhyped these associations as a marketing push.

[Arete]

23andme seems to be the first place to go. I have ordered from them. Hopefully they will provide me with the raw microarrays.

 

Then if someone discovers are particular snp linked to a disease, inorder for me to test if it is in my dna, I first I need to have the loci microarray of where it would be and then need the software search for it.

 

For example I would like to be able to test to see if I have a certain SNP that a researcher has linked to a disease. ( given that 23andme doesn't already test for it ).

 

a) 23andme sequences using an Illumina BeadChip array - https://www.23andme.com/more/genotyping/ They claim a million SNPs, but that doesn't match with any of Illumina's off the shelf chips, so either they've done some judicious rounding, or they use a custom chip.

 

b) If someone, in the future discovers a point mutation associated with a disease that is not present on the chip used by 23andme, you won't have that information. In order to future-proof your data like that, you'd need a whole genome analysis. The reason microarrays exist is because generating such an analysis is time consuming, expensive and requires specialist knowledge and high powered computational resources.

 

Thus, to generate useful, but efficient analyses that can scale up, certain studies subsample the genome using microarrays. This is the approach 23andme takes. It's a compromise - as producing an exhaustive, whole genome variation map with high confidence is a non-trivial task. Ergo, in answer to your original question, there is no easy way to deliver a confident, exhaustive variable call analysis of a whole human genome.There are however compromises like the microarray offered by 23andme which if you're a layperson without specialized skills in bioinformatics would appear to be adequate for most conceivable applications.

[Tridimity]

Might also be worth taking a look at Illumina's Every Genome:

 

www.everygenome.com

 

Still costs several thousands of pounds/dollars per genome... going to wait until the service is more affordable and see how my ACTGs line up. Happy sequencing