Solexa: advantages/drawbacks/comparisions

[bioinf]

Hi everyone. I'll be presenting the "Illumina GA I" in class as a seminar subject. I've already prepared the slides, but I lack a bit of information about it. I'd appreciate if you could give me some info on the following questions.

1) What are the major advantages of Solexa(GA1) sequencing machine compared to the other second-generation machines?

2) What are the major depicts of Solexa(GA1) sequencing machine compared to the other second-generation machines?

3) What would be the major change(s)/improvement(s) in GA IIx compared to GA I?

Please don't understand me incorrectly, I don't ask about something I'd just put to slides without understanding, but I'd rather need your thoughts as thoughts of people who used these machines or maybe some pointers from you to the information concerning the above-mentioned questions. Anything related to they subject is appreciated! Thx in advance!

[bioinfosm]

most of those are already discussed in this forum, like changes from GA I to IIx, even Illumina website might have those details; or try searching online for Solexa presentations as a lot of people put these numbers in their slides..

[drio]

@bioinf Perhaps you can read those threads and create an entry in the wiki summing up what
you find.

[volks]

the GA1 is from 2006/7, so the improvements to newer illumina machines should be quite obvious from just looking at the specs.

if you want to present the molecular mechanisms behind it, these pages have some nice illustrations:

http://seq.molbiol.ru/sch_lib_mp.html

http://seq.molbiol.ru/sch_clon_ampl.html

http://seq.molbiol.ru/sch_seq_polymerase.html

[bioinf]

Thanx, but I did prepare the presentation of the molecular mechanisms already. What is left is only this part about the advantages and depicts.
I'm not a biologist though, but rather a computer scientist doing the bio-informatics course.

[volks]

i dare to say that there are no/little advantages (except cost of the machine and having already established the workflow) of the GAI over the GAII(x), although i'm not an expert on illumina machines. same for the GAIIx compared to the HiSeq2000.

what you could compare nicely would be illumina to ABi SOLiD, 454, complete genomics, pac bio, ion torrent and so on.

[NextGenSeq]

The problem is that the instruments and software are upgraded so frequently that by the time any compilation is published it's out of date.

[clivey]

[QUOTE=bioinf;30246]Hi everyone. I'll be presenting the "Illumina GA I" in class as a seminar subject. I've already prepared the slides, but I lack a bit of information about it. I'd appreciate if you could give me some info on the following questions.

1) What are the major advantages of Solexa(GA1) sequencing machine compared to the other second-generation machines?

2) What are the major depicts of Solexa(GA1) sequencing machine compared to the other second-generation machines?

3) What would be the major change(s)/improvement(s) in GA IIx compared to GA I?

Of course the GAI, (or 1G Genetic Analyzer as it was also known) is now very old, dating from mid-2006 (early access) to late 2007. Any advantages have changed over time so may no longer be relevant. At launch the only competitor was the 454 system. Whilst this has longer reads at around 300bp (under a distribution) it was only about 3X cheaper per base than a capillary sequencer, it did have output in the 10s of megabase range. The 454 system depended on emulsion PCR and required ancillary equipment to support this. The workflow was relatively long winded and fiddly, so even though the run time of the machine was shorter, throughput was difficult to realise because of the workflow. Sales volumes were in the 10s for this system in the first couple of years. There were reported issues with homopolymer runs.

The GA had much shorter reads at around 25-35 bp (Q20 readlength), but the system produced over 1 Gigabase of data in under 3 days. The costs were about 100X better than capillary. The workflow, with the aid of a cluster station, was also relatively easy compared to emulsion PCR. The data quality could be high, and no homopolymer issues. However, data quality could be variable partly due to instrument variability, manual set up variability and some issues on reagent quality control (sometimes). Something like 200 systems were sold in the first year with over 40 orders in the first couple of months.

In the GAIIs, a better camera was used with other improvements to the optics allowing higher throughput (cluster density). Some of the issues around manual set up and reagent handling were improved lessening but not removing variability in customers hands. Later improvements to the de-blocking chemistry coupled with optical and protocol changes allowed longer reads at Q20 quality going beyond 50bp. However a lot of the run variability remained and new sources of error were introduced such as manufacturing faults in the flowcell channels and surfaces. It should be noted that run times lengthened with the GAII/x, exacerbated by the more widespread introduction of the paired end module (additional bit of hardware and chemistry) allowing paired reads to be done automatically, you could now do, say, 2X50bp in 8 days or there abouts with output in the 10s of gigabase range.

there were other tweaks not mentioned here.

The SOLiD systems appeared in 2007. You can look up the chemistry but these were ligation based sequencers working on beads with emulsion PCR like 454. The system however had higher output and lower costs (note Solexa and SOLiD systems were ~2X the price of 454 to buy). It gets contentious to discuss pros and cons around this system but initially it delivered 25bp reads using a color encoding scheme that complicated direct comparisons of data quality. suffice to say enough data of enough quality was produced by this system to allow resequencing experiments to be done cost effectively. However, this system also had a fiddly workflow, reliability issues on the instrument, and little post run analysis support. Yet, initially, it managed to scale its output somewhat faster than the GAs - the GA series later caught up and over took it in the form of HiSeq.

[bioinf]

Thank you very much guys. The information is helpful!
Thx clivey for a detailed answer!