PE sequencing of a lib with ONE end high and the other low complexity

[ein_io]

I am preparing a custom library for paired end sequencing.

It is a low complexity lib. To circumvent the problem with cluster calling I have included NNNNN from the end that will be sequenced first (right after the ACACTCTTTCCCTACACGACGCTCTTCCGATCT which I assume is used for the first read)

I wonder whether I need to put NNNNN on the other side as well? Does the reverse read use the cluster positional information generated during the first read?

any thoughts - deeply appreciated

[HESmith]

Try searching the forum; this issue has been addressed multiple times.

[ein_io]

tried searching the forum - found it mentioned once (in a thread discussing the bareback "delayed" cluster calling one guy says that clusters are not called during the reverse PE run) but not discussed explicitly

if you know the answer - can you tell me?

[pmiguel]

I am preparing a custom library for paired end sequencing.

It is a low complexity lib. To circumvent the problem with cluster calling I have included NNNNN from the end that will be sequenced first (right after the ACACTCTTTCCCTACACGACGCTCTTCCGATCT which I assume is used for the first read)

I wonder whether I need to put NNNNN on the other side as well? Does the reverse read use the cluster positional information generated during the first read?

Yes, the reverse read uses the cluster positions generated during the first read.

That said, I have noticed focus issues that seem to spring up when the instrument attempts a scan with no clusters fluorescing. (During the indexing cycles in my case.) My concern would be this would be especially an issue during the first few cycles of a read. Recent firmware upgrades have ameliorated this issue. But it still makes me nervous. But I think the instrument we have, the HiScanSQ, may be more susceptible to this issue than the more common HiSeqs.

Sacrificing those first 5 bases of your reverse read: I can't say it is worth it, but maybe.

--
Phillip

[ein_io]

Thank you very much, Phillip