Tip:
Highlight text to annotate it
X
All our impression of the performance of the Ion Proton, this is stable. It’s an easy
instrument to work with and it’s now we have been using it for a couple of months
and that has improved during this time, all the time. So now with the P1 Version 2 chip
we usually get like 80-85 percent loading efficiency and we get like 80 million reads
out from each. So this corresponds from an exome run usually ten GB of sequence data,
which is enough good actually to be able to put two exomes on one P1 chip. And we do we
have four units in our lab that we are operating in parallel and it seems like it doesn’t
matter which of the units we are running. It’s really that it seems like it performs
similar between the different units. So the homopolymers is a known issue with
the Ion sequencing technology, but we have seen that this has really improved with the
latest versions of the sequencing chemistry and also with the later version of that Ion
Torrent Suite. So we think that this is more or less overcome now and we think that the
quality of the exome data that we get out from the Ion Proton is good. We are using
the latest version of the Ion Torrent Suite, the version 3.6 in the relaxed mode with mapping.
We have found around 90,000 SNPs in each sample in the latest data sets and then 93 percent
of these SNPs are all found in DB SNP. They have also seen that ratio of false positive
SNPs in the Ion Proton data is significantly lower than comparing with other NGS platforms.
Talking about coverage of Exome sequencing data with one sample on one P1 chip we currently
have like more than 50x, quite the even coverage over the whole Exome. With the experience
that we had from our lab I think that everyone with normal skills from the lab can manage
to make a sequencing library if you follow the protocol carefully and if you evaluate
the different QC steps in the protocol carefully. I think that is really good with the scalability
on both PGM and Proton. You have the different sizes of the chip, so you can choose when
you really need/don’t need that much sequencing data. So when you have bar coding you can
put many samples on, but still it’s okay with the short-run time and also for the exome
or for a whole genome, you can be able to perform it in the sequencing part in less
than one day. So I think that’s important. Our ambition as a service provider is to choose
the most appropriate platform for a project, so I think the Ion Proton will be the choice
of project for exome sequencing with the current throughput. But in the future we really look
forward to run human genome sequencing on the P-2 chip and also be able to barcode several
exome samples on the P-2 chip to be able to keep the cost as low as possible for exome
sequencing. At the most recent exome sequencing project
we have been doing we have been working with a rare dominant disease where we had four
affected family members. There were two brothers and one child each for these brothers and
the sequence strategy was to make TargetSeq Exome capture and they had one sample per
P1 chip. And we got between 76-90 million reads out and 97 percent of these reads met.
And we did find around 90,000 SNPs in each sample and 93 percent of these SNPs were found
in dbSNP from before. And we found around 5K of InDels and 50 of these have been found
before. And the result is that 15 SNPs were found
in these samples and 8 of them are novel and 7 of them are known before. We have ended
up with quite a few candidates, so it will be straightforward work to validate these
samples. And I think that the performance looks nice, okay there’s a lot of data,
most of it maps and most of the majority of the SNPs are knowns from before, is in dbSNP.
So concluding we decided that this is a nice data set from these samples.