Transcriptome assembly: Why S-shaped length distribution?

[Jonathan87]

Hi,

when you (picture-)google for "transcriptome assembly length distribution" you find distributions that contain a little peak at the end.

For example here:

http://openi.nlm.nih.gov/imgs/512/82/3325976/3325976_pone.0035152.g002.png

What does this peak refer to? I cannot find any discussion results for that peak, maybe I use the wrong keywords.

[Brian Bushnell]

That's not a peak; everything above 1500bp is binned together. If you extended the maximum bin size to a very high number it would disappear.

[bastianwur]

Yup, and in case this doesn't apply: Most likely genomic contamination, which just lead to longer assemblies.