Why is a seventh base needed with MiSeq 6-base indices?

[StopCodon]

The MiSeq sample prep guide indicates that a seventh base is needed after the 6th base. What is the reason for that?

[mastal]

Possibly because with the Illumina technology the accuracy of the base calling for any particular cycle (having to do with things like phasing/prephasing) depends partly on the next cycle, so I guess if you want to have 6-bases for the index read, you would actually do 7 cycles.

[GenoMax]

@mastal is correct. Since phasing information can't be determined for the last base it is not used during (n-1 rule) downstream processing (demultiplexing).

[kmcarr]

@mastal is correct. Since phasing information can't be determined for the last base it is not used during (n-1 rule) downstream processing (demultiplexing).

This is true for the HiSeq, but the MiSeq will determine how many Index cycles to sequence based on the index length recorded in the sample sheet, without adding the the extra cycle. For standard TruSeq indexes (6bp) the MiSeq will only collect 6 cycles of index sequence and use all of them for demultiplexing. There is no way to tell the MiSeq to run and extra index cycle but ignore the last base when demultiplexing.

[GenoMax]

@kmcarr: You are right.

I tend to forget that we never use the on-board analysis on our MiSeq's. As a result we need to consider the n-1 rule while processing data externally.