We have pooled library and do de-multiplex to get fastq for each sample. However, out of these samples, only one sample shows overall low quality at the end of ends. Any possible reason?
how poor is "poor"? Can you post Screenshots so that we can compare a "good" and a "bad" sample? What sequencing platform was used? How was the library prep was performed? What kind of indexes did you use?
I am not sure the exact amount of samples. Probably around 10 samples were pooled in the same chip. For one sample, the last 10 bases got quality lower than 30 from fastqc.
It was amplicon-sequencing on Hi-Seq.
This could simply be a bad library that has relatively short inserts which is causing adapter sequence read-through. This generally leads to a rapid drop in Q scores since you start getting low nucleotide diversity.
Thanks. This is a possible reason. I will check the insert size of this sample.
We found that 'N' is enriched at the end of the reads. If this is poor template generation due to low nucleotide diversity? Will the sequence enrich with 'N' or some other phenomenon?
Most likely. When the basecaller loses confidence to call a base it is going to put N in that spot.
BTW: Is the sequence before N's any good or are these just very short inserts or primer dimers? That entire sample may have failed.