I am not sure if I should mention the name of the company that did the exome scan.
The American government appears to want to shut down the entire genomics industry.
It makes no sense! Very soon everyone will have their genome sequenced. Genomics will
be a new economic pillar. I searched all over the internet and I could not find a commercially
reasonably priced American exome sequencer (Thankfully, someone suggested the company that did the scan).
The exome scan that was performed yielded almost 8 Gigabase pairs, with almost 4 Gigabase
pairs in the exome. This translated into almost 65x times coverage of the exome (62.1 Megabase pairs).
Surprisingly, of the 62.1 Megabase pairs, there were only 60,000 variants. This illustrates that mutations in the exome have been selected against during evolution (as would be expected).
The summary report for the exome scan noted that over 20 stop loss mutations, almost 70 stop gain, over 90
frameshift insertions and over 60 frameshift deletions were found. The report did not disclose the identity of
these mutations, though these will be a priority for me to investigate further.
This scan was actually quite pricey. The current $1000 genome scan from Illumina yields 100 Gigabase pairs. So,
1 Gigabase pair should cost about $10. An 8 Gigabase pair exome sequence (assuming the genome cost point) should
cost $80. The final base pair price for the exome scan was much closer to 10x that of the genome scan.
I started off and wanted to be cheap and go for 30x coverage. However, I did not fully appreciated that 30x coverage
means, on average, a base pair would be sequenced 30 times. I did some reading on the internet and found that a research grade exome is considered to have 30-50x coverage, though a clinical grade exome is considered to have 70-100x coverage. That is only an average! In the 65x coverage exome scan that was performed for me, some of the base pairs in the exome scan were sequenced over 200 times, many were not sequenced at all.
The fixed cost of just getting the exome scan up and running is about $350. From there on out the marginal cost for 1x coverage is about $5. When I realized all this I moved up my coverage request to 60x. (I actually finally got 65x).
I am so glad that I spent the extra $150. At only 30x, you might only have 70-75% of the SNPs being considered high quality. At 100x, 90-95% might be considered high quality. My scan yielded almost 91% Q30 quality SNPs (Q30 means an error rate of 10 e-(Q/10), that is Q30 is an error rate of 10 e-(30/10) = 10e-3 =.1%. A Q30 SNP has 99.9% probability of being correct). Each scan is different, though I am happy now for the choice that I made. Paying another $200 for the full
100x scan would not have likely yielded information worth the marginal cost.
It would be very sensible for me now to do a full genome imputation from the results of my exome scan. Such a scan
should give me a very accurate full genome result, and cost almost nothing.
It is too bad that technology has not been developed that would allow uniform coverage across the exome.
Even at 100x coverage (average), 5-10% of the SNPs might not be considered high quality. However, a SNP that has been called 30 times is considered high quality. If the sequencers could simultaneously align the SNPs and target regions with
low coverage then you could have a higher quality sequence result with less sequencing. This would cut the cost of sequencing considerably. It appears that the sequencers do not respond to the sequencing results in real time.
The company wanted to stay clearly within the FDA legal guidance concerning no medical interpretation of the genome.
Even still, I regret not asking the company if they could enrich the sequencing for known Alzheimer disease exon regions.
It might be possible to enrich the probes for certain specified areas in an exome.
I read an article that mentioned that approximately 50% of base pairs in an exome scan are so called off-target.
Surprisingly, many of SNPs that have been found to be associated with diseases (including AD) are "off-target"
(i.e. outside the exome). Ordinarily, these off-target base pairs would simply be discarded. I asked the exome sequencing company to provide me with these off-target base pairs. They agreed. I have not been able to determine
which file contains these SNPs, though I expect that such a file might contain about 40 Megabase pairs of these
off-target base pairs (of the 4 Gigabase pairs of raw off-target sequence).
I had to beg the company to take my money. They needed to have a company meeting to decide
what to do. Apparently, they had very little experience with the direct to consumer market. They did
not stock saliva collection kits, so I had to ask around to buy one. The first company (the leader in spit kit technology) refused to sell me the kit! (They required that I disclose the identity of the company doing the exome sequence). Fortunately, one company agreed to sell me a kit.
Shipment to America was very reasonably priced and without customs declarations hassles. Postage only cost $3! If the collection tube had been narrower, postage would have only cost $1! With 23andme, postage cost more
than $50 and there was an extensive customs declaration form that needed to be completed.
The basespace account has several analysis tools available. I am having some trouble getting the apps to function properly. The company that did the scan is still listed as owning the scan. This appears to be preventing me from running the app. Being an early adopter requires doing extra leg worka and putting up with hassles. For the benefit of human civilization, one must struggle through.
I have downloaded GenomeBrowser and it looks as though it will be helpful.