Library failure - any help appreciated

[firefly2280]

Hi all

I am having trouble getting a MiSeq run to work and was hoping someone might be able to help

I have swab samples from surfaces in a hospital, i have DNA extracted using a bead beat and Qiagen column. Then I have pooled the extracted DNA from bed spaces and PCR's using 16S primers and custom barcodes for my library prep.

I have qPCR'd almost all of the samples individually before pool with a Staph generic assay and there are a few with not much in and some that may have been inhibited but I would think that would be diluted out with the pooling.

So, I have PCR'd to add barcodes, run a size select gel to purify - seen bands on all lanes. I have Qubited, adjusted to 4nM, pooked 96 samples, qubited again and then used this as my library. Denatured and prepared as per Illumina protocol and the samples are not working.

Does anyone have any ideas? thanks

[Heisman]

What do you mean by not working? How do you know they failed?

[firefly2280]

What do you mean by not working? How do you know they failed?

The Phi X has clustered but the run then fails, so no reads at all from my samples.

[kwaraska]

I would confirm the sequences you put on with the barcodes.

"Old style"-non Truseq single end adapters don't work on the MiSeq although they do work on the HiSeq.

[firefly2280]

I would confirm the sequences you put on with the barcodes.

"Old style"-non Truseq single end adapters don't work on the MiSeq although they do work on the HiSeq.

Thanks, they are the exact same ones as being used by other MiSeq users in my lab and they are getting good results

[GW_OK]

Can you post SAV specs? Cluster density, %PF, base intensity plots....

[firefly2280]

Can you post SAV specs? Cluster density, %PF, base intensity plots....

Hi, sadly not as the runs failed. the samples failed to cluster, nothing worked. apart from the Phi X.

[GenoMax]

Hi, sadly not as the runs failed. the samples failed to cluster, nothing worked. apart from the Phi X.

If phiX worked (I assume that means you got the sequence) then you should have the graphs/stats that GW_OK asked you to post. They will only show phiX data.

I have come across only one instance of a sample where it would not work on a HiSeq after multiple runs but did work on a MiSeq. Perhaps if you have access to a HiSeq you could try running your samples there (if the libraries are compatible).

[firefly2280]

If phiX worked (I assume that means you got the sequence) then you should have the graphs/stats that GW_OK asked you to post. They will only show phiX data.

I have come across only one instance of a sample where it would not work on a HiSeq after multiple runs but did work on a MiSeq. Perhaps if you have access to a HiSeq you could try running your samples there (if the libraries are compatible).

ok, thanks will do later. I dont have anything with me. Im really new to this (you can probably tell)!

[Number6]

If you have a custom primer for your sample, the MiSeq requires a higher Tm than primers on the HiSeq (I believe this is discussed in other threads in this forum). We had a sample that only sequenced on the MiSeq when we used a longer sequencing primer, and failed with a shorter version of the same primer. Might be worth checking the Tm (if applicable).

[firefly2280]

If you have a custom primer for your sample, the MiSeq requires a higher Tm than primers on the HiSeq (I believe this is discussed in other threads in this forum). We had a sample that only sequenced on the MiSeq when we used a longer sequencing primer, and failed with a shorter version of the same primer. Might be worth checking the Tm (if applicable).

Many thanks Number6. The primers are all being used successfully by a colleague, so we dont think they are the issue. Thanks though