Index Registration Failure?

[puggie]

I was reading the ScriptSeq Index PCR Primers (1-12) protocol, and it says:

Since the Index sequence of pooled, multiplexed ScriptSeq libraries will be read simultaneously, it is important to maintain color balance for each base of the Index sequences in the pooled library. Otherwise, Index sequencing will fail due to registration failure.

Then I'm thinking how will this affect the sequencing? Could the registration failure lead to an incorrect barcode assignment? Or will the read just be assigned some nonsense sequence?

Also why is it important that the short barcodes match up for the red and green laser? What about the entire read length, shouldn't they also "match" up then?

The reason I ask is that i have RNA seq data from a six-pooled Illumina run using indexes that apparently should not have be mixed

Any inputs are appreciated.

[NextGenSeq]

This is more critical if you use 5' barcoding since the clusters are mapped after the first half a dozen or so bases. With 6 barcodes you probably have enough signal from each base that it will be okay.

If it's 3' barcoding it probably won't matter at all. You can usually figure out what barcode it is even if there is an error at one position of the barcode.

[puggie]

Okay thanks for your reply. The case is I'm doing paired-end sequencing, the barcode is in the 5'-end adaptor. Know perhaps as a newbie I don't understand the principles behind the technology fully (will attend a Illumina seminar on RNA seq later this month).

However, I have received this paired-end data, upon doing boxplot of sequencing statistics I saw that for the reverse read (R2) the sequencing quality dropped significantly after 40 bp (the reads are 101 bp total). I was told that they experienced some kind of technical fault at the sequencing facility. For the forward read (R1), the quality was okay dropping only slightly beyond 80 bp as I understand is normal.

Now, Im having difficulties understanding if this could influence the proper barcoding of the six pooled libraries that was run on the flowlane? Could reads have been mixed up? Of course the sequencing will be re-run, however now Im looking at preliminary data, and want to acertain that forward/reverse reads may not have been mixed up.