High percentage of Off-target reads

[finswimmer]

Hello,
we are using TruSight Capture Kits from Illumina for Targeted Sequencing.

Normally we have around 25% off-target reads within one run. But from time to time there are runs with 75-85% off-target reads resulting in low coverage in the target region.

What can be the reason for that?

Our guess is, that it has something to do with the hybridization or the washing step afterwards. Maybe the washing soluting wasn't on room temperature? Maybe the hybridization time was to long (but we never exceeded the 20h recommended in the manual)? If the hybridization time was to long, than in the first or in the second hybridization?

Have you had similiar problems and found out what the reason was? Any other guess?

fin swimmer

[husamia]

it depends what the rest of 75-85% is. Try to size select before sequencing, this gets ride of small fragments <150 that may be lost due to adapter contamination

[finswimmer]

Hello husamia,

we always check the fragment size of our library with a Bioanalyzer. There the library show a normal distribution of the fragment sizes. Also cluster density, cluster passing filter, Q Values, percentage of total aligned reads, the distribution of samples within the library looks fine.

There are only two values that are abnormal:
- the mentioned high percentage of off-target reads
- the concentration of the library after preperation. It's up to 3-4 times higher than normally. That's why we guess it has something to do with the capture or washing step. Either the hybridization probes captured to much or within the washing step to much unspecific region stay in the sample.

So maybe it's possible to check the concentration with Qubit after washing and if it's to high (but what is "to high"?) we can repeat one more wash?

fin swimmer