The Spike-In Cooperative (SIC)
The S.I.C. is a group of like-minded investigators who sometimes need to borrow a little bit of sequencing, and who are sometimes willing to lend a little sequencing. The SIC is made possible by the DNA Sequencing Innovation Lab at the Center for Genome Sciences & Systems Biology (CGS&SB).
The short version:
- Make a library using unique 9bp indexes assigned to your lab.
- Submit the library to Jess and ML at the CGS&SB to spike-into a sequencing run.
- You will receive demultiplexed fastq files of your spike-in sample.
- Cost? $175, for about 500,000 reads.
The long version (FAQ):
How does it work?
Sometimes you need a small number of reads to see if a new experiment is working before you commit thousands of dollars to a bigger experiment. As a member of the SIC, for a modest fee, you can have your samples “spiked-in” to someone else’s run. You, the “guest” will get about 0.5 million reads out of the “host’s” reads.
You turn your samples in to the CGS&SB on the 4th floor of the McKinley Research Building, along with the typical information about the sample type and indices (SubmissionSheet). The CGS&SB takes care of all sample handling and logistics. You get the data back.
What type of run will we be spiked into?
Most samples are spiked-in to a 2 x 250 or 2 x 150 run.
What kind of libraries can I spike in?
Anything with your unique indexes, and Illumina paired-end adapter sequences.
How do we tell our samples apart?
By the indexes. Every lab in the SIC will order their own unique 9-bp indexed library primers (see table link at bottom of this message). Therefore, anyone’s samples will be able to be spike’d into anyone else’s and they can all be sorted out post-hoc. This allows for a lot of logistical flexibility and hopefully faster turn around.
What does it cost?
You pay a modest fee of $175 to the CGS&SB for the sample handling and insurance, and they take care for the rest, including the splitting the reads by the indexes (“demuxing”). The only other cost associated is in ordering your own 9-bp indexed primers. These will cost you about $10 a primer from IDT. They do not need special purification, unless they are over 80 bp total. In which case PAGE is recommended.
When would I want to be a ‘guest’ or spiker-or?
Whenever you just need a small number reads. Examples:
- You’ve developed a new capture protocol and you want to see the efficiency before scaling up
- You want to sequence a small genome or a plasmid library
- You’ve got samples you would like to QC before running your own full run
Why would I want to be a ‘host’ or spike-ee?
Why not? If you already spike-in 10% of PhiX into your run, this would take the place of up to 8% the PhiX. You don’t loose any of the reads associated with your sample. CGS&SB takes care of all of the logistics. The only thing you do differently is to get in the habit of using your labs unique indices. And there is an insurance policy to cover your run if something goes wrong (see below). So it costs you nothing, but helps your colleagues in the collaborative.
Plus, to be a member of the collaborative (a guest), you have to be willing to be a host as well. That’s the deal.
You are permitted to occasionally ‘opt-out’ a particular run as ineligible for hosting a spike-in sample. (For example if your library itself is experimental and you are worried that it may fail for some reason).
What are the risks?
The major risk is that guest library takes a large proportion of the host’s lane due to mistakes in mixing the sample (over-spike). To ameliorate this risk, part of the fee goes into an insurance pool. If the host gets less than 85% of their expected reads due to the spiked-in sample, then the host will get their entire lane re-run (with no spike), paid from this pool.
Several labs have been spiking into their own lanes for some time and rarely have any problems. Nonetheless, by signing on to the collaborative, you are saying that you accept these risks, and that anything greater than 85% of a lane is acceptable yield when you are the ‘host.’
For the guest, we will be returning approximately 250,000-750,000 reads. As the ‘guest’ libraries are often expected to be experimental in nature, there is currently not an automatic guarantee for the ‘guest’ for read number.
How long will it take to get my library spiked into a lane on the sequencer?
If there is good participation, there will be many host lanes available, and the queue turn-around will be relatively rapid. Queues will be managed by CGS&SB.
If there are ever a sufficient number of ‘guests’ (spike-ors) in the queue waiting (about 6-10 libraries depending on the run type), the CGS&SB will simply run an “all spike” lane, and everybody gets extra data for free. Thus the program is also a program for lane-sharing, without you having to do the logistics of finding nine other people to split the lane with. Overall there is a for a relatively fast turn-around time (less than two weeks).
What library QC is part of the $175 fee?
A BioAnalyzer chip will be done to QC the sample. Any issues detected will be discussed with the investigator.
What are the deliverables?
On all runs, you will get your standard demultiplexed fastq file(s) back from CGS&SB.
If requested, you will also have access to a summary data of the ‘demuxing’ of the whole lane, for troubleshooting purposes, as well as meta-information about the libraries your lane was shared with.
Who’s got what primers? How do I get my unique indexes?
Click here SIC_ indexes.xlsx to see everyone’s unique indexes, or to pick some of your own. There are currently a few pre-designed primers available to pick from here. All have a hamming distance of at least two from any others, or any known GTAC, Illumina, or other commonly used indexes. Detailed instructions are in the file. After picking, email the file back to us and we’ll update it here.
Any questions not answered here? Or, want to claim some primers for your lab?