Protocol version 1.9 and oligonucleotide design sheets

[Joe Foley]

The current version of the Smart-3SEQ protocol, 1.9, is attached to this message. Also included are several files that can be used to simplify ordering the custom oligonucleotides from IDT, depending on the indexing strategy.

Common files:

• Smart-3SEQ oligos for IDT bulk input.ods: Sequences for the 1S and 2S primers, universal P5 PCR primer, and 48 legacy-indexed P7 PCR primers. The contents of this file can be copied and pasted into IDT's "Bulk Input" interface for ordering custom single-stranded DNA oligos, though the 1S primer should be ordered as a DNA oligo, the 2S as an RNA oligo, and the PCR primers as "Ultramer" DNA oligos. Note the highlighted indexes have problems described in the protocol.
• indexing_comparison.pdf: Diagram of the different indexing strategies.
  

Legacy 48-plex single-indexing:

• Legacy single-index PCR primers for IDT plate.xls: Sequences and plate coordinates for only the universal P5 PCR primer and 48 legacy-indexed P7 PCR primers. This file can be uploaded into IDT's interface for ordering 96-well plates instead of single tubes, which is cheaper and possibly more manageable. Note the highlighted indexes have problems described in the protocol.

96-plex unique dual indexing:

• UDI PCR primers for IDT plate.xls: Sequences and plate coordinates for only the 96 unique-indexed PCR primers (P5 and P7 on separate plates). This file can be uploaded into IDT's interface for ordering 96-well plates.
• index_usage_table.pdf: Simple chart you can print and hang up on your freezer to tally which indexes you've used, in order to balance your usage and avoid running out of some of them before others.
  

Regardless of indexing strategy, you always need the 1S and 2S primers from the ODS file, and those must be prepared with RNase-free HPLC. To use legacy single-indexing, you need the universal P5 PCR primer and any number of legacy-indexed P7 PCR primers. For dual indexing, you need any number of pairs of P5 and P7 PCR primers (they are designed to be used together, i.e. index P5-01 goes with P7-01, P5-02 with P7-02, etc.). For most practical purposes, e.g. tube labeling, we find it is easier to refer to the 96-plex indexes by their plate coordinates (A01, B07, C11, etc.) than by their original ID numbers (01, 50, 83, etc.).

Between developing the legacy single-indexing PCR primers and the new UDI primers, IDT advised us we could reduce costs by switching to their "Ultramer" synthesis for the PCR primers, which eliminates the need for HPLC purification, and by using only one phosphorothioate (between the final C and T). This change is reflected in all the attached files, but for the record we have never actually ordered Ultramer single-thiol versions of the legacy single-index primers or HPLC double-thiol versions of the UDI primers. RNase-free HPLC is always required for the 1S and 2S primers.

We have also never ordered the 96-plex P7 primers, because our lab currently uses an unusual i5-only 96-plex strategy: we use all 96 sequences of the i5 PCR primers with a universal P7 primer (like the indexed one in the protocol except no index sequence). This requires unusual sequencing conditions but gets better results than i7-only indexing and is cheaper than UDIs. See the "PCR primers for higher-plexity pooling" section of the protocol.

EDIT 2020-02-03: slightly changed the names of the oligos in the IDT ordering sheets to be more consistent with one another and the new dilution template; also increased the default synthesis scale of the universal P5 so there's enough to make a full batch of dilutions at the suggested volumes.

[Joe Foley]

Here is the design for the universal P7 PCR primer that can be combined with the 96-plex P5 PCR primers from the UDI set to do i5-only indexing.

EDIT: the 4 nmol scale won't give enough yield for very many uses of 96 aliquots, so this file is corrected to order the 20 nmol scale. Copy and paste the entire row from the spreadsheet to IDT's "Bulk Input" mode to get the recommended options.

[jm googlecalendar]

These resources are just sooo useful for those of us starting out, so many thanks yet again. 

As I am beginning this journey from scratch, I hope to fully emulate the i5 indexing strategy your describe so lucidly.  I am wondering, exactly, the source of the i5 index sequences listed in the UDI PCR primers for IDT plate spreadsheet files, which catalog the full-length 96 i5 indices (TruSeq PCR P5-01--> TruSeq PCR P5-96), ready for uploading to IDT.  As best I can tell, they are not congruent (sense or antisense) with any IDT UDI adapters currently listed on the IDT website (e.g., IDT xgen-udi-umi-indexes.xlsx, or idt8-indexes 384.xlsx ).  The only reason I ask is my indexing needs are for 128 distinct libraries, and I had it in my head that I might order a 384 well plate's worth of i5 indices on P5 primers and not need to worry about mixing, or at least order the 32 extra i5 indices.  However, I got a little nervous about typing in just any old index sequences in the middle of the TrueSeq P5 primer sequence, and I have looked rather extensively for a file online find none.  This suggests to me that you have again done the heavy lifting on the 96 i5s, though if you know of a electronic list of more than 96 i5 TruSeq i5s I would value knowing where I might download it.

Ordering  a custom plate of 200 pmoles per oligo from IDT is rather plenty for my experimental work (some 20-40 reactions worth), which gave me reason to think about ordering 128 oligos with i5s more given the price (CAD~$1500 for 96 wells) of Ultramer purity DNA oligos.

In the meantime, I understand that our sequencing facility runs NovaSeq equipment, which has two flow-cells per run, hence, I could use 96 indices per flow cell and have plenty as long as my accounting and pipeting is careful.

Just want to be sure my strategy makes sense while avoiding silly mistakes, or repeating mistakes others have already solved.

Many thanks for any thoughts you may have.

j

[Joe Foley]

The index sequences in the provided IDT ordering sheets are the same as the ones in Illumina's kits. I don't know which kit is which anymore because they keep changing every couple of years, but pragmatically you can see the current and past indexing schemes in the IEM configuration files. You can also find Illumina's 384-plex UDI list in there.

However, are you sure you need to do batches of 128 libraries routinely? It's easier to work with batches of 96 or less because most instruments are designed around the 96-well plate format. It might take extra sequencing runs to accommodate all your libraries but you offset that cost not just with fewer indexed PCR primers to purchase but also fewer to measure and dilute, which can take a lot of time when there are a lot of primers.

--
You received this message because you are subscribed to the Google Groups "Smart-3SEQ" group.
To unsubscribe from this group and stop receiving emails from it, send an email to smart-3seq+...@googlegroups.com.
To view this discussion on the web visit https://groups.google.com/d/msgid/smart-3seq/415b2d48-ba34-44c4-82dc-c6509ff404e2n%40googlegroups.com.

[jm googlecalendar]

I had, naively, thought to use one i5 index per sample (in my experiment) which has 128 biological replicates.  The calculus I used was simply based on the cost per oligo at the 200 pmol scale (~CAD$16/oligo) vs. the cost of a sequencing run at our university (~$3000 for NextSeq; ~$4000 for NovSeq).  For the 32 "extra" indices beyond your 96, the cost is ~$500. 

The practical reality will be that I will be processing perhaps only a dozen or two LCM caps at a time, and I expected to process them as quickly as possible to get them to a safe (i.e., cDNA) state for storage before pooling for a sequencing run.  Practical experience might drive a different solution (e.g., storing the caps @ -80C), but I had thought to make a single run per sample, repeating for different cell types (128 * 4) for our analyses. 

On a practical note, I am trying to imagine how best to store and use the 200 pmol plate to:  (a) achieve the proper 4uM concentration (dry vs. diluted), (b) minimize freeze-thaw (aliquot or not), (c) minimize contamination (plates vs. tubes), and (d) keep track of everything (your freezer sheet is nice). 

Any suggestions would be most useful.

Thanks again.

j

[Joe Foley]

OK, if I'm understanding correctly, what you're planning is a total of 512 libraries and you have access to either a NextSeq 500/550 (not 1000/2000?) or a NovaSeq. Assuming your samples are from an organism whose transcriptome is similar in complexity to human, eight flow cells of NextSeq High Output or four of NovaSeq SP with the shortest (cheapest) available single-end reads should be enough for that many samples from LCM (see figure S31 in the paper). Unlike the NextSeq, the NovaSeq lets you load different samples in two lanes (NovaSeq Xp workflow), so you could just reuse 64 indexes eight times on either platform. If you scale up to the higher-throughput NovaSeq flow cells, then you could do fewer runs with more indexes. You can check the pricing on that but be sure you're looking at the cheapest single-end read lengths because long reads aren't especially helpful in Smart-3SEQ, particularly with LCM material (figure S30). You might also want to look for a iSeq/MiniSeq/MiSeq to check the library quality as cheaply as possible before scaling up to one of the expensive platforms. And search this Google Group for discussion about the NovaSeq; our lab is only experienced with the NextSeq but others have reported success on the NovaSeq, though maybe not yet with LCM samples.

We routinely store our LCM caps at -80 C for a while before doing a library prep. Our main time-sensitive practice is to do the LCM quickly after staining/deparaffinating the slides.

We also store our original "plate" of primers from IDT (it's more like an array of fancy tubes) at -80, after remeasuring all of them on the spectrophotometer and preparing another set of standardized dilutions of the individual primers at the same concentration (e.g. 50 uM), which we also store at -80 except when we take them out to make small (~20 uL) working dilutions of the PCR-ready primer pairs, which we then store at -20. Everything after the original "plate" is in individual tubes since it's hard to seal a real plate securely enough for long-term storage and you usually only need to take out a subset of primer pairs for any given experiment, but for ease of pulling out the tubes we split the 96 of them into two racks of 48, every other row.

To view this discussion on the web visit https://groups.google.com/d/msgid/smart-3seq/5f935feb-9620-40f1-9ba2-f360d4af4927n%40googlegroups.com.

[jm googlecalendar]

Thanks for your very practical and pragmatic advice; much appreciated.  Sometimes (most times) the elegant science stumbles through tripwires of niggly details that are crucial to getting the job done.

[jm googlecalendar]

Before we order our plate of DNA indexed oligos from IDT, I am interested in your thoughts about the "Ship Option" for these "plates" (I see they are available in 500 uL Matrix screwtop tubes [ThermoScientific]).  The IDT menu gives two choices:  (a) dry (lyophilized--as specified in the Smart-3SEQ protocol), and (b) wet.  Choosing the "wet" option opens menus for various "IDTE buffers" (TE buffer 10 mM Tris/0.1 mM EDTA) at pH 8 or 7.5 (your openwetware.org notes suggest pH 8.0), AND various concentrations (2-10 uM in 1 uM increments). 

Given your experience and recommendation for quantifying individual oligomers, and the incredibly valuable spreadsheet for calculating concentrations for each specific i5 oligo, I am wondering if "factory diluted" oligos were available to you previously (i.e., were lyophilized oligos were the only option?) and if you had any faith in the "factory" calculation of concentration (we intend to check!). 

The IDT website specifies:

Every Ultramer DNA Oligo you receive will be deprotected and desalted to remove small molecule impurities. In addition, your oligos will be sequence-verified via proprietary ESI-mass spectrometry methods*, and quantitated twice by UV spectrophotometry for accurate yield measurements.

If ordering, I would ask for the 10 uM and not depend on IDT to hit the necessary 4 uM concentration for the Smart-3SEQ.

Many thanks for your thoughts as we prepare to open the chequebook.

j

[Joe Foley]

Years ago we ordered the factory-diluted oligos @ 100 uM in TE pH 8, because they were the same price as lyophilized and at least it saved one step, but we still remeasured all of them on the Nanodrop and made a whole batch of 96 redilutions at exactly 50 uM according to the UV spec. At the time the numbers were messy but they didn't do UV spectrophotometry for us, so if they offer that now then it might save even more trouble. However, it's still a good idea to trust but verify, i.e. at least check that all the oligos you receive are close to the expected concentration even if you don't expect to redilute them. If e.g. the caps aren't tight during shipping there could be differential volume loss to evaporation (I've heard of this recently). I would definitely order at least 10 uM concentration in case you do need to adjust some of them.

To view this discussion on the web visit https://groups.google.com/d/msgid/smart-3seq/2d6c9166-6b37-4400-9033-7a1b311f7a2en%40googlegroups.com.

[jm googlecalendar]

Thanks for confirming.  Here we go...:)