Low cluster generation and biphasic reads quality!

[melNGS]

Hi all,

I had some issues with my last run and I hope someone can help me to find explanations. We loaded a library of 10 pm from a starting dilution of 8nM, following the Illumina guideline, on the miSeq and I had a cluster of 337 K/mm2 (really low!). Also in the end of the run the quality of the reads dropped from 98% to 66% showing two groups of reads with two different qualities. I attached the pic from the end of the run, so you can have a look of what I'm talking about!

Can DNA overloading influence negatively the generation of the clusters? Is it possible that Q score decreased as the majority of fragments are less than 150 bp?

Thanks for your suggestions,

Melissa

[share]

Dear Melissa
Check the thumbnail images for over clustering. If there are excess clusters MCS can't resolve them properly.
Check phasing/pre-phasing values and error rate pattern any sudden dips and bumps??

Were these libraries QPCR quantified??

Thanks

[GenoMax]

Hi all,

Can DNA overloading influence negatively the generation of the clusters? Is it possible that Q score decreased as the majority of fragments are less than 150 bp?

Thanks for your suggestions,

Melissa

There can be a negative impact on quality/yield of overloading. If the inserts are smaller than the run length then one can potentaily start reading into the "oligo lawn" on the flowcell and this will lead to dramatic drops in quality towards the ends of reads.

Can you post an image of the heat map of Q-scores? Were these sequences expected to have low-complexity?

There are several threads on this forum that address problems/observation when running low-nucleotide complexity libraries on MiSeq.

[mcnelson.phd]

Can you post a screen shot from SAV of the % base plot? As Geno said, reading into the anchor oligos will cause quality to drop, but you can tell if that's occurring because you'll see a characteristic increase in A bases in the % base plot.

As for the cluster density, what type of libraries were you sequencing? I've seen a few issues with the newest version of RTA where it will call a low cluster density even though the images look like the cluster density should be higher. Our FAS could never fully explain why, but we've only seen it two times and that was when the libraries themselves had issues as well.

[melNGS]

Thanks for your answer.

I've looked the thumbails and it seems there is an uneven distribution of clusters! Attached there are some pics, and in one of them I can see a blob of clusters (am I right?).

The library was prepered with Haloplex and quantified by bioanalyser.

For that run we decided to generate only fastq files because our intention was to analyse with CLC bio software afterward; do you know a way to re analyse with miSeq Reporter in order to have all the run information like phasing/pre-phasing values, heat map of Q-score, etc, % of bases,..

[AStretton]

The values will have already been calculated for phasing/pre-phasing you just need to look at the data in Sequence Analysis Viewer (SAV) which should be on the MiSeq desktop. It's a really useful way to have a look at your data quality.

I have had problems sequencing fragments less than 150 bases but it seemed to be the library causing the problem and I ended up redesigning the primers to make the fragments slightly longer. It was the cheaper option.

[melNGS]

The % phasing/prephasing are:
Read 1: 0/0.190
Read 2 (I): 0/0.52
Read3: 0.8/0.61

There is an increasing of errors in the prephasing but I cannot see a drammatic problem!
However the qscore heat map does look problematic....

What concern the % of the bases, I can see an increase of C bases, so, for what I understand, it didn't read the "oligo lawn"!

[GenoMax]

Are these metagenomic (or low nucleotide diversity) samples?

You may have short(er) inserts than you expected. Have you tried to analyze the data at all?

Are you running the latest version of MCS and RTA on this MiSeq?

[melNGS]

We looked the bioanalizer report and it appeared a high concentration of adapters. Apparently it is a common thing using Haloplex from Agilent as library preparation kit. We are trying to change some variables and run again. Hopefully it will work better this time!

After trimming the extra adapters, the data looks fine!

Thanks all so help me out!