Cost Comparison Between Microarray And Rna-Seq Experiment

Question

[polarise]

Hi,

What is the typical cost of carrying out a microarray experiment compared to a similar experiment using RNA-seq? For example: 10 human exon arrays versus RNA-seq for the same 10 individuals.

For microarray: cost of reagents, cost of 10 exon arrays; assume the chip scanner is present in the lab. For RNA-seq: cost of reagents, cost of sequencing at a commercial sequencing center.

I do not have any wet-lab experience so please advise how to better phrase this question.

Thanks!

PK ^_^

Answer 1

[Obi Griffith]

This is very difficult to answer as prices will vary widely depending on academic vs non-academic pricing and the exact nature of the service. Even working inside a genome center it can be very hard to get solid estimates of what an "RNA-seq experiment" really costs internally. If you have a nearby genome center or core services at your university it is possible that you will be able to get a better deal through some kind of collaborative arrangement. On the microarray side, prices are probably more standardized but will still vary widely depending on which platform you are talking about. On the RNA-seq side it will really depend on exactly what kind of sequencing and how much you do.

To give you a general idea, LC sciences currently advertises $2500/sample for library preparation and sequencing for part of a HiSeq lane giving you up to 60 million 2x100bp reads of RNA-seq data. I think that is a typical price and amount of data you might expect from a commercial vendor. Depending on what you are trying to accomplish with the RNAseq experiment you may want more coverage than that. I list LC Sciences only because they were the only service provider that I found which actually listed prices online. It is generally the case that prices are only available upon request and tend to be negotiable.

There are many providers for Affy exon array services online. It seems the average price is around $800 per sample for full processing if you include the cost of array itself. But, there are also lots of options here (different labeling strategies, amplification, etc) that can increase costs. If you have your own hyb station and scanner for the array then maybe you can get it done for about half of that per sample.

[polarise]

Fabulous response. I now have good ball-park figures in mind. Thanks. ^-^ PK

[Obi Griffith]

This answer is now quite out of date with respect to RNAseq costs. These still vary widely depending on the protocol, target number of reads, etc. However to update the ballpark numbers I estimate that a typical RNA experiment starting from isolated RNA (including sample management, Illumina TruSeq Stranded library construction, and sequencing) might cost $600, $800, $1000 or $1400 per sample if multiplexing 8, 4, 3 or 2 samples per lane on a HiSeq2500/HiSeq4000, with 2x100(or 125)bp reads for a modest sized project (~20 samples). Costs will be slightly lower on a HiSeq4000 and will likely go lower yet once HiSeqX is released for RNAseq. NOTES: Remember these are rough estimates. Your local core may (likely) charge a little more. Some genome centers may also be able to do better for larger projects. These prices don't include any indirect costs that may come into play. The amount of multiplexing you plan for will depend on the analysis/experimental goals of the project. Adding an exome capture step may allow increased multiplexing, again depending on your experimental goals.