Amplicon sequencing: Where do short reads come from

[TimK]

Hi,

have worked with a lot of shot gun metagenomic data but have recently looked into a couple of 16S/18S amplicon sequencing projects. One thing that immediately caught my eye was that if I order shot gun metagenome Illumina reads 2x250 then I only get reads of 250bp length. However, the amplicon sequencing results I've seen included read from 35-250bp (pre-QC). And even after joining the reads there are still joined fragments of e.g. 60-400bp length.

This raises two (main) questions for me:
1. How is it possible that I get reads that are less than 250bp in a 2x250bp run? Are those PCR artifacts or does the sequencing cycle stop for some cells?
2. How can I get joined fragments that are shorter than 250bp (which I would assume is the minimum possible length for 250bp paired end reads and even then they'd have to overlap 100%)? The only explanation I'd have would be chimeras ...

Any suggestions?

[nucacidhunter]

1- Sequencing centre must be trimming adapters or low quality bases. once a cluster passes filtre the instrument will collect data for all 2x 250 cycles.

2- You can get shorter joined fragments if your insert is short provided that adapters has been trimmed.