cluster density variation

[HeinKey]

Since a few weeks we occasionally observe very big variation between raw cluster densities on our Single read flow cells.
Even if we load multiple lanes from the same library tube we sometimes see this. eg. variation was between 730K/mm2 and 164K/mm2 loaded from the same tube!!

Also aliquoted Phi-X for a dedicated control lanes sometimes has as little as 20k/mm2 clusters where it should be 500k/mm2.
aliquoting is something we do for years now with reproducible results up until a few weeks ago.
We only observe this with single read flow cells. We never see this with PE flow cells.

We used different sequencers, fluidics on cBot and sequencers are always checked and OK. Still we observe this, and can't get it under control.
Illumina is not aware of a SR flow cell issue, and also has no clue how to control this.

Has anybody seen this as well, or has anybody a clue what could be causing cluster density variation?

[IDDan]

We saw wild cluster density variation on MiSeq until Illlumina instructed us (with a new method) to carryover less NaOH.

Our experience may not be applicable to you. If NaOH carryover is not applicable, check pH of your NaOH...assuming NaOH is used for HiSeq???

[Anguy]

Hi HeinKey,

Are these variations being observed on the same flowcell but different lanes? We occasionally load multiple lanes with same library from the same dilution and it should always perform quiet constantly. One of my guesses is when there are smears of dirty spots on the flowcell it can severely damage the machine's ability to calculate viable clusters. We have seen that on GA when oil was not loaded well (or sometimes the oil stick on thermal station and create smears on flowcell as the run progress).

[HeinKey]

We saw wild cluster density variation on MiSeq until Illlumina instructed us (with a new method) to carryover less NaOH.

Our experience may not be applicable to you. If NaOH carryover is not applicable, check pH of your NaOH...assuming NaOH is used for HiSeq???

IDDan; thx for your response.
We see cluster density variation within a single flow cell. Variation between lanes loaded from the same diluted single stranded library. Mixed well!
If NaOH is accidized I would expect to see only variation between individually denaturated libraries.

[HeinKey]

Hi HeinKey,

Are these variations being observed on the same flowcell but different lanes? We occasionally load multiple lanes with same library from the same dilution and it should always perform quiet constantly. One of my guesses is when there are smears of dirty spots on the flowcell it can severely damage the machine's ability to calculate viable clusters. We have seen that on GA when oil was not loaded well (or sometimes the oil stick on thermal station and create smears on flowcell as the run progress).

Hi Anguy,
Thnx for your response.
I agree with you that local disturbances would result in cluster density variation. I checked images and cluster density plots. This all looks normal with equally spread clusters all over the lanes, just in much lower densities. so I still doubt if the Single Read flow cells were OK here. Are the grafted oligo's in lower concentration, or damaged? and why only in Single Read flow cells?