Bench philosophy: A practical guide to Next Generation Sequencing platforms

Nine Question Method
by Karl Gruber, Labtimes 05/2011

Chuck Berry, in an earlier generation, once wrote a song called the 13 Question Method that offers some guidance on finding the way to a girl’s heart. Today Karl Gruber asks nine questions that may directly lead you to decide, which Next Generation Sequencing platform to set your heart on.

I remember when I first learned to sequence DNA on a good old ABI 377 sequencer. Operating the machine was quite tricky: you had to make your own polyacrylamide gel and load the amplified DNA on a vertical slab-gel, before pushing the “start” button. A sequencing run took seven hours and delivered 36 to 48 samples (depending on the comb), with about 700 base pairs.

Fast and accurate

Today, pretty much every bit of the ABI 377 technology is history. Current Next Generation Sequencing (NGS) technologies offer much faster, much more accurate and efficient sequencing. It’s not the intention of this article to dwell on the complex chemistry and technologies behind each NGS platform. It should rather provide a clear and straightforward guide to help choose one of the existing NGS approaches that best suits the potential user or curious researcher’s needs. The article is mostly targeted at someone who wants to use these technologies, e.g. by hiring a private company or a local facility that will analyse their samples. If, on the other hand, you are thinking of purchasing NGS equipment to do everything yourself, I recommend you to look over the review by Travis Glenn (Field Guide to Next Generation DNA Sequencers. Molecular Ecology Resources., 2011, doi: 10.1111/j.1755-0998.2011.03024.x). Ralph Schlapbach from the Functional Genomics Centre Zurich provides a well-illustrated explanation of each NGS platform (

Early attempts

The very first NGS technology emerged around 1990. It was developed by Lynx Therapeutics, and was named “Massively Parallel Signature Sequencing” or MPSS. As often happens with new technologies, this first prototype was a bit too complicated and never made it to the outside world, however, it got things rolling, so to speak. Lynx Therapeutics was eventually bought by a larger company and their MPSS technology was mostly discarded. But one remaining aspect of those early NGS technologies is the production of millions of small DNA fragments, which finally need to be assembled.

The first commercial sequencer using a single-molecule approach was built by the manufacturer HeliScope. Unfortunately, these instruments are no longer sold but sequencing service is still provided by this company. Currently, the most popular NGS platforms are: Roche’s 454 sequencing system, Illumina’s HiSeq and Genome Analyzer machines and Life Technologies’ SOLiD system.

Sequencing with a 454 system is based on beads that carry the template DNA for the emulsion PCR (emPCR) reaction. Each DNA bead is deposited into a separate well of a pico titer plate during the sequencing reaction. Hence millions of sequences are obtained.

Illumina’s technology uses a solid glass surface to produce clusters of identical DNA molecules via Bridge PCR (primers are attached to the solid surface).

The name of Applied Biosystems (Life Technologies) platform SOLiD stands for Sequencing by Oligonucleotide Ligation and Detection. Here, billions of short sequences (between 35 and 75 bp) are produced, using an emPCR method similar to 454 sequencing.

All these technologies give great results but there are some general guidelines, depending on the applications. Genome Analyzer and SOLiD produce a massive amount of data. Hence they are, e.g., better qualified to identify polymorphisms then the 454 system. However, if you want to assemble or build a draft genome sequence from a not-previously-sequenced species, then the 454 system is a better choice because of its longer reads.

Basic parameters and main advantages/ disadvantages of the most popular NGS platforms
Platforms Millions of reads per run Bases per read Primary errors Error rate Main Advantages Main Disadvantages
454 GS Jr.
0,1400Indel0,0Long Reads,
Low cost per run
High cost per Mb sequenced,
fewer reads than Illumina
454 FLX
1400Indel0,0Long ReadsHigh cost per Mb sequenced,
fewer reads than Illumina
454 FLX+1700Indel0,0Even longer readsHigh cost per Mb sequenced,
fewer reads than Illumina
3,4150 + 150Substitution≥ 0.1%Lower error rate,
paired reads
High cost per Mb sequenced,
shorter reads than 454
250100 + 100Substitution≥ 0.1%Lower error rate,
paired reads
High cost per Mb sequenced,
shorter reads than 454
320150 + 150Substitution≥ 0.1%Lower error rate,
paired reads
High cost per Mb sequenced,
shorter reads than 454
HiSeq 1000
500100 + 100Substitution≥ 0.1%Lower error rate,
paired reads
High cost per Mb sequenced,
shorter reads than 454
HiSeq 2000
1000100 + 100Substitution≥ 0.1%Lower error rate,
paired reads
High cost per Mb sequenced,
shorter reads than 454
Hard to compare

It is quite hard to compare these technologies, as there are no tests out there to match different NGS platforms. So, one must rely on whatever information is available from each company’s website. This approach is also subject to problems, as we don’t know in detail, how each company has optimised the sequencing runs to obtain the best results. However, we can still gain a fair idea about main characteristics, to determine a set of pros and cons. The table at the bottom of the previous page gives a summary of the main advantages and disadvantages of the most popular NGS platforms (adapted from Glenn, 2011).

Choosing a suitable NGS platform

But which NGS platform do you really need? Here are some pointers. Well, the first question, is of course: what is your scientific question? The second would be: how large is your budget? If you can answer these two questions you are half way to getting your NGS data. Below are further questions and guidelines that you may check before buying a certain NGS platform or ordering a custom-sequencing service that uses a particular NGS technology.

  1. What organism are you going to work with?
    Human. Go to 2.
    Non-Human. Go to question 3.

  2. OK, this was easy. Your best bet would be to use companies specialised in working with human data, such as Complete Genomics, deCode genetics or Knome.

  3. What type of data do you want to get?
    Genomic DNA. Go to question 4.
    RNA, aka Transcriptome data. Go to question 5.

  4. Is there a reference sequence for your target species, e.g. has this species been sequenced and annotated already?
    Yes. Go to 11.
    No. Go to question 8.

  5. Is there a reference sequence for your target species?
    Yes. Go to 6.
    No. Go to 7.

  6. Illumina HiSeq and SOLiD 5500 are the favourites, followed by HeliScope.

  7. For de novo transcriptomes, 454 FLX+ and Roche´s 454 FLX+ and Illumina HiSeq take the prize. Illumina has quite a bit of a lead by producing almost 500 times more reads (1 mil. per run for 454, vs. 500-1,000 mil. for Illumina).

  8. Is it a plant or animal genome you want?
    Go to 9.
    Or is it a BAC, plastid or microbial genome?
    Go to 10.

  9. Illumina HiSeq seems to be your platform.

  10. Roche’s 454 FLX+ seems like your best match, ideally using their pooling option, with tagged samples.
    This way, you can pool together multiple individuals of your population and run them together in one reaction. This will be the most efficient use of NGS technologies for your question.

  11. Good, having a reference genome helps a lot.
    Are you re-sequencing a genome?
    Yes. Go to 12.
    No. Go to 13.

  12. There is a match between Illumina HiSeq and SOLiD 5500. Your choice.

  13. Are you targeting certain loci?
    Go to 14.
    Or are you analysing transcripts?
    Go to 15.

  14. Illumina HiSeq is your weapon of choice. Second runners include SOLiD 5500 and Illumina MiSeq.

  15. Illumina HiSeq and SOLiD 5500 take the lead here, followed by HeliScope.

Important! Please notice that the final choice of platform is exclusively based on the scoring guide provided by Travis Glenn ( You should also check the most recent data on all these platforms, since the manufacturers are constantly updating and improving both chemistry and equipment.

Last Changed: 10.11.2012