understanding multiplex sequencing on GAIIx

[seqgirl123]

I am trying to understand how multiplex sequencing on GAIIx works andrunning into some questions:

How come Rd1 SP can't just sequence the entire first strand, including index, to produce the complementary strand to which Rd2 SP will anneal to?
As it is from Illumina's illustration, Rd1 SP generates the complementary strand of the template, the complmentary strand then gets removed, Index SP anneals and reads the index, then the complementary strand is regenerated so Rd2 SP can anneal and sequence it. Why does the complementary strand have to be made twice? Why can't Rd1 SP just generate the complementary strand, including index? Meaning, what stops it so that a separate Index SP has to be used to read through the index?

What happens during cluster amplification because the illustration only displays the sequencing aspect?

I also saw somewhere Illumina has two approaches it uses to incorporate the index read, I think this was during the sample prep part. One can use a 2-step procedure to implement the barcode into the template, or a 1-step procedure. I cannot find that diagram again, can someone attach it here? I do remember the first approach had the barcode sequence in front of the template (5' end I think), and the second approach had the barcode sequence behind the template (3'end I think) after sample prep completed. What is the reasoning behind the two approaches? For sequencing, it looks like approach 2 during sample prep is most commonly used since the barcode is behind the template, not in front of it. But if approach 1 was done how would the sequencing process differ? Would only one Rd SP be used?

[Heisman]

I am trying to understand how multiplex sequencing on GAIIx works andrunning into some questions:

How come Rd1 SP can't just sequence the entire first strand, including index, to produce the complementary strand to which Rd2 SP will anneal to?
As it is from Illumina's illustration, Rd1 SP generates the complementary strand of the template, the complmentary strand then gets removed, Index SP anneals and reads the index, then the complementary strand is regenerated so Rd2 SP can anneal and sequence it. Why does the complementary strand have to be made twice? Why can't Rd1 SP just generate the complementary strand, including index? Meaning, what stops it so that a separate Index SP has to be used to read through the index?

If I'm understanding your question correctly then the answer might be that if you were to sequence the index on the first run, you would also have to sequence all of the bases of the sequencing primer. In addition, you would have to sequence all of the bases of the insert. So instead of a 42 bp read, for example, if you have an insert of 200 bp you would have to sequence the 200 base pairs, then the sequencing primer, then the index. That simply can't be done now.

What happens during cluster amplification because the illustration only displays the sequencing aspect?

When a library is loaded on a flow cell, it contains your fragments of interest (inserts) that have the PCR primers on the end. During cluster amplification one of the fragments has one of its ends hybridize to a complementary adapter on the flow cell, and PCR is done on that spot to get many more of the same strands.

I also saw somewhere Illumina has two approaches it uses to incorporate the index read, I think this was during the sample prep part. One can use a 2-step procedure to implement the barcode into the template, or a 1-step procedure. I cannot find that diagram again, can someone attach it here? I do remember the first approach had the barcode sequence in front of the template (5' end I think), and the second approach had the barcode sequence behind the template (3'end I think) after sample prep completed. What is the reasoning behind the two approaches? For sequencing, it looks like approach 2 during sample prep is most commonly used since the barcode is behind the template, not in front of it. But if approach 1 was done how would the sequencing process differ? Would only one Rd SP be used?

Indexing is shown in the diagram. With barcoding, the adapters that are ligated on contain an extra few base pairs that are used to identify the sample. So, with barcoding, you have [insert][barcode][sequencing primer][flow cell adapter], while with indexing you have [insert][sequencing primer][index][flow cell adapter]; both of these are only taking one side of the insert into account. The downside to barcoding is that you waste a few sequencing cycles to actually sequence the barcode, and in addition these are the first few cycles which are generally the highest quality (lowest error rate). Also, if you sequence barcoded samples you may run into trouble with the image software distinguishing between different clusters. The benefit to barcoding is that you can order custom adapters to include the barcode and then proceed with a normal non-indexing protocol.

Hope that helps!

[kmcarr]

I am trying to understand how multiplex sequencing on GAIIx works andrunning into some questions:

How come Rd1 SP can't just sequence the entire first strand, including index, to produce the complementary strand to which Rd2 SP will anneal to?

The nascent strand generated from the Rd1 SP is not covalently anchored to the flow cell. Once you separate it from the template strand used for Rd1 synthesis, which would be required to use it as a template, this newly synthesized strand would simply wash away.

As it is from Illumina's illustration, Rd1 SP generates the complementary strand of the template, the complmentary strand then gets removed, Index SP anneals and reads the index, then the complementary strand is regenerated so Rd2 SP can anneal and sequence it. Why does the complementary strand have to be made twice? Why can't Rd1 SP just generate the complementary strand, including index? Meaning, what stops it so that a separate Index SP has to be used to read through the index?

The illustration is horribly non-representative of relative sizes of the primers, index and insert. For a current paired-end library the insert may be ~400bp (if you are doing 2x150 that is). To read all the way through this, even at HiSeq speeds would be > 2 weeks.