ABySS and Velvet "number of contigs and contig length in pair-end and single-end datasets"

Question

[Abdul Rafay Khan]

I have run Velvet and ABySS using default parameters on my pair-end and single-end datasets, I have two questions related to the generated contigs.

1. Why Velvet has excess number of contigs in pair-end case compared to single-end?

2. Why the N50 contig length of single-end reads using ABySS is so short?

[Sej Modha]

There could be a number of different things that might have led to this results such as k-mer size for the assembly, quality of the single vs paired-end data and a bunch of other parameters that could have been used to build the assemblies.

[benv]

I am not familiar with Velvet usage, but if you provide the abyss-pe command and verbose log output (add "v=-v" to abyss-pe command), I could have a look. Output of abyss-fac (N50 etc.) is helpful as well.

It is standard practice to run assemblies for a range of k-mer sizes to achieve the best possible results.