Using custom MID labeled primers in conjunction with TruSeq kit for multiplexing

[GMDickson]

I am using the same primer pair to amplify the same target sequence in 384 different individuals. One thing I should have mentioned is that I am amplifying an MHC exon which can have up to 8 different alleles per individual. This makes genotyping challenging since you are never really sure how many alleles an individual should have. I used a 2 step PCR approach in the past, but found that this introduced a high level of PCR errors. This probably wouldn't have been a problem for a sequence without repeats, but for an MHC gene it made it impossible to genotype a large number of my samples. It is for this reason that I am considering the TruSeq LT kit for library preparation, which uses ligation rather than PCR to attach adapters. I am trying to keep costs down, so I am trying to avoid any protocols that require unique primer combinations for each individual. I also have heard that sequence diversity is not as big an issue as it has been in the past. For my previous run, we spiked in a high concentration of PhiX and were fine.

Hi jpleonard2000,

Did you ever develop a protocol to do this (using ligation to add the adaptors rather than a PCR)?

Thanks.

[GenoMax]

I typed a long reply to Ion user but the DDOS protection system ate it.

Grab a copy of your response via clipboard before hitting submit. Lesson learned the hard way.

[kmcarr]

I typed a long reply to Ion user but the DDOS protection system ate it.

When I have some time I'll re-submit.

[Ion user]

I would not do it this way, far too complex. Our lab regularly multiplexes up to 384 samples in a single MiSeq using two step PCR. The first PCR reaction amplifies your target(s) using target specific primers with universal oligo sequences at their 5' ends. PCR #2 uses Illumina compatible, dual indexed primers which target the universal oligos now at the ends of your primary amplicons. The second PCR is only 4-6 cycles. There are 24 barcodes for the forward primer and 16 for the reverse which provides 384 unique pairs. This set of 40 primers can then be used to make indexed libraries for any target, all you need is to order a single pair of target specific primers with the universal oligos at their 5' ends for each target you want to sequence.

In our lab we use the Fluidigm CS primer design and ordered our set of index Illumina primers, but you could just as well order Nextera index primers from Illumina (they now have 384 index kits) and follow their design guidelines for your target specific primers. Adapt the 16S primers in this document to fit your needs.

Diversity issues with regard to sequencing amplicon libraries on a MiSeq has not been a significant issue for a while, since Illumina made software improvements to deal with it. For amplicons we typically spike in ~5-7% PhiX and this works well most of the time.

Hi,kmcarr:
Did you try this like that， put 40 primers in one tube instead of two step PCR? I tried,but there were many small fragments such as primer dimmers or other non-specific fragments and I didn‘t size selection successfully with Ampure beads？

[jpleonard2000]

I am using the same primer pair to amplify the same target sequence in 384 different individuals. One thing I should have mentioned is that I am amplifying an MHC exon which can have up to 8 different alleles per individual. This makes genotyping challenging since you are never really sure how many alleles an individual should have. I used a 2 step PCR approach in the past, but found that this introduced a high level of PCR errors. This probably wouldn't have been a problem for a sequence without repeats, but for an MHC gene it made it impossible to genotype a large number of my samples. It is for this reason that I am considering the TruSeq LT kit for library preparation, which uses ligation rather than PCR to attach adapters. I am trying to keep costs down, so I am trying to avoid any protocols that require unique primer combinations for each individual. I also have heard that sequence diversity is not as big an issue as it has been in the past. For my previous run, we spiked in a high concentration of PhiX and were fine.

[kmcarr]

I would like to sequence 384 small amplicon samples on a single MiSeq run. ………

Has anyone used a similar approach to multiplex a large number of samples? Thanks.

I would not do it this way, far too complex. Our lab regularly multiplexes up to 384 samples in a single MiSeq using two step PCR. The first PCR reaction amplifies your target(s) using target specific primers with universal oligo sequences at their 5' ends. PCR #2 uses Illumina compatible, dual indexed primers which target the universal oligos now at the ends of your primary amplicons. The second PCR is only 4-6 cycles. There are 24 barcodes for the forward primer and 16 for the reverse which provides 384 unique pairs. This set of 40 primers can then be used to make indexed libraries for any target, all you need is to order a single pair of target specific primers with the universal oligos at their 5' ends for each target you want to sequence.

In our lab we use the Fluidigm CS primer design and ordered our set of index Illumina primers, but you could just as well order Nextera index primers from Illumina (they now have 384 index kits) and follow their design guidelines for your target specific primers. Adapt the 16S primers in this document to fit your needs.

Diversity issues with regard to sequencing amplicon libraries on a MiSeq has not been a significant issue for a while, since Illumina made software improvements to deal with it. For amplicons we typically spike in ~5-7% PhiX and this works well most of the time.

[nucacidhunter]

Since the reads would begin with the Roche MID sequences, would this introduce enough sequence complexity to prevent run failure? Also, what about including amplicons produced with regular untagged primers? These would be shorter than those amplified with the MID labeled primers, but they would be easy to identify simply due to their lack of a MID sequence. Can anyone foresee any problems with this approach? Has anyone used a similar approach to multiplex a large number of samples? Thanks.

You have not given enough information on your proposed workflow so it is not easy to comment or foresee potential issues. It is possible to combine inline barcodes introduced to 5' end of primers with Illumina's adapter index. To increase diversity inline barcodes (MID) should be variable length 5-8 base long and carefully designed to balance colour during sequencing. Questions are:
1- Are amplicons targeting same or similar sequence in different samples?
2- Are you going to ligate adapters to amplicons?

[jpleonard2000]

I would like to sequence 384 small amplicon samples on a single MiSeq run. The amplicons are 238 bp without primers. I had an idea of using primers tagged with 10bp MIDs available from Roche, and then using the TruSeq LT kit to provide an additional level of multiplexing. The TruSeq LT kit provides unique adapter combinations for up to 24 samples, so I would need 16 unique MID-labeled primer combinations to allow multiplexing of 384 samples. So, I could then order 4 unique forward primers and 4 unique reverse primers.

The delivered reads would be demultiplexed according to the TruSeq adapters, but I would then have to demultiplex each of these "samples" further according to my own MID-labeled primers.

Since the reads would begin with the Roche MID sequences, would this introduce enough sequence complexity to prevent run failure? Also, what about including amplicons produced with regular untagged primers? These would be shorter than those amplified with the MID labeled primers, but they would be easy to identify simply due to their lack of a MID sequence. Can anyone foresee any problems with this approach? Has anyone used a similar approach to multiplex a large number of samples? Thanks.