[jcosta]

Great

This is a great idea.
Soon I will get my hands on Ion Torrent. I would gladly contribute with data/runs from this technology.
I remember the early days of microarrays when I was part of a team dealing with the MIAME guidelines. This will surely be of great value for everyone.

Cheers

[pareje]

If we are going to do this then the possibility of a very interesting project arises. We have put a lot of effort at CRI into experimental deisgn to reduce the impact of confounding factors see http://1.usa.gov/jh7U1h "The cost of reducing starting RNA quantity for Illumina BeadArrays: a bead-level dilution experiment."

As a group of people running well over 1000 instruments (from the google map) we could put together a definitive MAQC style paper as a group collaboration.

This would also allow comparison of just about every technology. As a group we have access to everything!

Design might look like:

• Genome - pick one relativley simple such as C. elegans
• Prep - ideally several groups would make libraries from a defined strain from http://www.cbs.umn.edu/CGC/index.html
• Insert size - we could vary this by collecting 200, 300 and 400bp insert libraries
• Barcode - use Illumina TruSeq barcodes on libraries so a pool could be made for that platform. Similarly for other platforms.
• Instrument - use any instrument. Perhaps a different person could co-ordinate each technology separately, I'd be happy to organise Illumina.
• Run Type - SE and PE of different lengths. If barcoded then this could be spiked into any run pretty easily.
• Institution - run the pool across as many institutes as possible
• Other - there are so many factors we could vary with the right design and co-ordination. Feel free to suggest others.

Bringing all this together with the bioinformatics community on SEQanswers doing the analysis would be great. The run requirements would be small so anyone with an instrument could get involved. Similarly for some of the basic analysis efforts. It would be a real community challenge. And we could make the whole thing public so anyone could get on board.

We'd end up with a rich data set that showed how similar the technologies were on a genome sequencing study. We'd be able to say a lot about how insert size and run type affect genome sequencing or analysis. And we would have a lot of instrument variation data, is SOLiD more variable than Illumina, Ion, PacBio?

Lastly if you wanted to make it a personalised Genome experiment we could offer our own genomes up for sequencing. I think I'd be happy to put mine forward. Anyone else?

James, this is a great idea. As we assume "crowd sequencing" would have great power, I would only add to your design two things: 1. a few more genomes, to make sure the differences are not C. elegans specific. and 2. a computational pipeline, to see how analyses influences actual results.

[james hadfield]

If we are going to do this then the possibility of a very interesting project arises. We have put a lot of effort at CRI into experimental deisgn to reduce the impact of confounding factors see http://1.usa.gov/jh7U1h "The cost of reducing starting RNA quantity for Illumina BeadArrays: a bead-level dilution experiment."

As a group of people running well over 1000 instruments (from the google map) we could put together a definitive MAQC style paper as a group collaboration.

This would also allow comparison of just about every technology. As a group we have access to everything!

Design might look like:

• Genome - pick one relativley simple such as C. elegans
• Prep - ideally several groups would make libraries from a defined strain from http://www.cbs.umn.edu/CGC/index.html
• Insert size - we could vary this by collecting 200, 300 and 400bp insert libraries
• Barcode - use Illumina TruSeq barcodes on libraries so a pool could be made for that platform. Similarly for other platforms.
• Instrument - use any instrument. Perhaps a different person could co-ordinate each technology separately, I'd be happy to organise Illumina.
• Run Type - SE and PE of different lengths. If barcoded then this could be spiked into any run pretty easily.
• Institution - run the pool across as many institutes as possible
• Other - there are so many factors we could vary with the right design and co-ordination. Feel free to suggest others.

Bringing all this together with the bioinformatics community on SEQanswers doing the analysis would be great. The run requirements would be small so anyone with an instrument could get involved. Similarly for some of the basic analysis efforts. It would be a real community challenge. And we could make the whole thing public so anyone could get on board.

We'd end up with a rich data set that showed how similar the technologies were on a genome sequencing study. We'd be able to say a lot about how insert size and run type affect genome sequencing or analysis. And we would have a lot of instrument variation data, is SOLiD more variable than Illumina, Ion, PacBio?

Lastly if you wanted to make it a personalised Genome experiment we could offer our own genomes up for sequencing. I think I'd be happy to put mine forward. Anyone else?

[Seqasaurus]

regarding the actual subject of the thread, does anyone know where to get $-750 ion torrent sequencing? (see table 2)

[DZhang]

Rather than depend on what shows up randomly in the published literature, could not this forum run a yearly "SEQanswers Open" on a standard template? This event could actually complement student lab training programs and promote platform understanding, excellence and standardized techniques among upcoming trainees.

I think it is a great idea. Through reading some next-gen-related blogs, I came across a google-shared document:

https://spreadsheets0.google.com/spr...Lw&hl=en#gid=0.

We do not have to use it as is but the idea is there. Can we adopt a similar document/format on this site? I believe the next-related software wiki took a similar route. At the beginning, it is a bunch of threads and later there were structured ways to present, update, and organize information.

Douglas

https://www.contigexpress.com

[Corey]

now that users can make barcoded libraries this could be done pretty easily. Someone just needs to make a big stock and send it out for free.

I concur, this is a great idea. A little confused though about the best way to standardize a sample so that everyone is sequencing the same material. Is the idea to grow up and do a big E. coli (e.g. DH10B) genomic prep, have one or a few sites sequence this material and settle on a reference? Alternatively, to ensure everyone is sequencing the same fragments from the same reference, we could clone a library of a modest genome (maybe yeast is better that E. coli so there will not be any contaminating host DNA) and then distribute that library as DNA to anyone who asks? I'd be willing to make the library if anyone thinks it worthwhile. Not sure that the advantages of the latter approach outweigh its negatives.

[krobison]

Neat idea! I'd suggest a standard bacterial strain (E.coli DH10B?) as then all variants can be considered errors.

A related notion would be a place for people to post any standard dataset or even just some standard run statistics for their platforms.

[ECO]

now that users can make barcoded libraries this could be done pretty easily. Someone just needs to make a big stock and send it out for free.

Good idea. I'd be up for it and or distributing/organizing...

[james hadfield]

now that users can make barcoded libraries this could be done pretty easily. Someone just needs to make a big stock and send it out for free.

[Joann]

Basically with reference to the literature above, groups (in training, even) could periodically run a standard template on their machine and properly report the data people want to see. Nothing fancy.

[Heisman]

This is interesting, but I'm not sure I get exactly what you mean. A yearly review of literature? A yearly benchmark of sequencer performance? A conference? Maybe this should be split out into a new thread?

Honestly I think only once a year is too few considering how rapidly things are moving and the number of emerging technologies. Maybe twice a year?

[ECO]

Rather than depend on what shows up randomly in the published literature, could not this forum run a yearly "SEQanswers Open" on a standard template? This event could actually complement student lab training programs and promote platform understanding, excellence and standardized techniques among upcoming trainees.

This is interesting, but I'm not sure I get exactly what you mean. A yearly review of literature? A yearly benchmark of sequencer performance? A conference? Maybe this should be split out into a new thread?

[Joann]

SEQanswers Open?

Rather than depend on what shows up randomly in the published literature, could not this forum run a yearly "SEQanswers Open" on a standard template? This event could actually complement student lab training programs and promote platform understanding, excellence and standardized techniques among upcoming trainees.

[BadDNA]

That study looks good (very detailed). It is exactly what is needed to compare across platforms; someone independently running the same templates on multiple platforms (as is called for within the Field Guide). Unfortunately, the supplementary files are messed up (the first file is doc & should be excel, and the second is excel, but should be doc), so I couldn't tell when the data were collected (i.e., what version of the chemistry for each platform was used?). From the read-lengths, it looks like the data are perhaps 3 years old. I'm not tossing any stones (I have much older unpublished data), but the chemistry & performance is a moving target. The 454 data are probably still pretty relevant (and surprisingly good), but I would guess the current SOLiD chemistry would do a little better & the current Illumina chemistry would do significantly better.

If there were more recently collected data analyzed in the same way, that would be very interesting to compare with these data.

[KerryOdair]

Comparison of Sequence Reads Obtained from Three Next-Generation Sequencing Platforms

Next-generation sequencing technologies enable the rapid cost-effective production of sequence data. To evaluate the performance of these sequencing technologies, investigation of the quality of sequence reads obtained from these methods is important. In this study, we analyzed the quality of sequence reads and SNP detection performance using three commercially available next-generation sequencers, i.e., Roche Genome Sequencer FLX System (FLX), Illumina Genome Analyzer (GA), and Applied Biosystems SOLiD system (SOLiD).

http://www.plosone.org/article/info%3Adoi%2F10.1371%2Fjournal.pone.0019534