Singletons in De novo assembly

[mandar.bobade60]

Hi,

I am doing de novo assembly for mitochondrial genome for that after being used Velvet and Amos(minimus2), I have got more of singletons than assembled sequences. So, around 5000 nucleotides come in fasta file and around 1100000 nucleotides are coming in singletons files. Still, I merged amos output singleton file and fasta file and generated assembly. Finally I used SSPACE and Gapfiller also. This has been done using paired end data without merging multiple fasta files for input.

My only concern is related to singletons, as they were high in numbers hence I couldn't exclude them. Also, adapter contamination was low, as apparent from Fastqc results. So, precisely my question is whether or not good to include singletons for de novo assembly.

[mandar.bobade60]

It would be great if someone directs me on this query..

[GenoMax]

Have you taken a sample of the singletons and checked to see what those are by blast. If they are not what you expect (non-mitochondrial contaminants) then you would be better off dropping them.