Quality drop in the Reverse read of paired end library (illumina Hiseq 1000)

[anurupa]

Hi
Recently we have prepared 3 paired end libraries (2X100) and sequenced on illumina hiseq 1000. we have checked the quality of the sequenced reads and found that for one library both forward and reverse reads showed good quality fatsQC reports. However for other 2 libraries the fastQC report for forward read (R1) is very good. But the fastQC report for reverse read is looking really bad (especially the per base quality score). I suppose if there is any problem with the library preparation then it should affect both the reads (R1, R2) which is not the case here.Could anyone let me know the reason why this is happening?


Thanks

[kohanlon]

I saw a similar problem with some of my libraries and after sending data to Illumina, found that my libraries were too concentrated. When the cluster density gets too high, the second read quality can be affected significantly. The instrument has trouble doing the turnaround when there are so many clusters so your Qscores drop off. You should be able to look at your images and see if you have a higher density of clusters at the front of the flow cell or at the back. If the clusters have a higher density at the back of the flow cell, it is an indication of over clustering and could be the cause of your drop in Qscores. I would try using a lower concentration of your library for the input and see if that clears up the problem.

[anurupa]

we have loaded three libraries in a single lane (multiplexed). As mentioned out of 3 libraries 1 library din't show any problem for forward(R1) and reverse(R2). we have checked the cluster density which is ~600 k/mm2. I think it is optimal value and the problem arises only if it is above 900 k/mm2