successful clinical sequencing

[niceday]

Just run a breast cancer sequencing sample on our PGM and got 15.2Mb.
Result.

[GW_OK]

But did it mean anything?

And how does the sequence impact the clinical side?

[flxlex]

Could you please (please?) share the read length distribution? I have asked Life for this and they just say '100 or 200bp'. Surely the real lengths are more variable?

[IonTorrent]

Attached is a read length distribution for a single 314 run of E coli DH10B. Insert size on the library was about 120 bp.

There's more information on this run available in our Spring 2011 Performance Technical Note - available at iontorrent.com or in ioncommunity.iontorrent.com.

Sincerely,

mike lelivelt
Director of Bioinformatics
Ion Torrent - Life Technologies

[epistatic]

If mean insert size was ~120 bases was the read trimmed to take off sequencing into the adapter or is that included in the read length distribution? I have been told quite tight size ranges for library insert sizes making the instrument less useful for doing QC on a Paired-end insert sized library.

[krobison]

Just run a breast cancer sequencing sample on our PGM and got 15.2Mb.
Result.

Which versions of the reagents & software were you using? How many reads in the dataset?

[NextGenSeq]

Okay so that is 0.5% of the human genome. I guess the personal genome machine will be sold to bacteria to sequence their personal genomes?

[hmartin]

Which versions of the reagents & software were you using? How many reads in the dataset?

We were using the old chemistry, what we are calling v1.2 (to match the s/ware version), and not the Xpress version that is out now. As to reads, we got 147508 with a mean length of 103bp.

[hmartin]

But did it mean anything?

And how does the sequence impact the clinical side?

It did indeed mean something. As a first pass, proof of principle, it exceeded our expectations. All SNPs were concordant with our CE results including a small (pathogenic) deletion detected by CE as a heterozygous frame-shift. As for clinical impact, cost is a major factor and the 314 chip is not competitive compared to our CE approach. However, the 316 chip becomes significantly cheaper than CE if multiplexing is exploited sufficiently. The new chemistry and OneTouch automation should lessen the hands-on time and we can see how the Xpress chemistry improvements should result in further data quality improvements and reduce library processing. The biggest hit is in data analysis speed, something that costs a significant amount of time for a clinical scientist with CE data.

[Fabian]

I'm guessing this was amplicon sequencing? What was the average coverage depth across the sample? I know the PGM isn't supposed to have stellar accuracy, so depth becomes king.

[GW_OK]

It did indeed mean something. As a first pass, proof of principle, it exceeded our expectations. All SNPs were concordant with our CE results including a small (pathogenic) deletion detected by CE as a heterozygous frame-shift. As for clinical impact, cost is a major factor and the 314 chip is not competitive compared to our CE approach. However, the 316 chip becomes significantly cheaper than CE if multiplexing is exploited sufficiently. The new chemistry and OneTouch automation should lessen the hands-on time and we can see how the Xpress chemistry improvements should result in further data quality improvements and reduce library processing. The biggest hit is in data analysis speed, something that costs a significant amount of time for a clinical scientist with CE data.

Very cool!

[jp.]

can some tell me what are the best read length , lib size, replicates and other standards for WGS PE Hiseq2000 of human samples. I am expecting complete best design without worrying about cost (I got funding) which will not make problem while analyzing data.
thank you in advance

[Elcannibal]

This thread is a great concept, just like sequencing a chunk of heterogeneous cancer cells.

[jp.]

any solution for my post please ?

http://seqanswers.com/forums/showthread.php?p=112098#post112098