Getting started, advice, etc

[JohnUrban]

Hi,

Thanks for the great tool.

I have messed around with smartdenovo on a few datasets using the following:

~/software/smartdenovo/smartdenovo/smartdenovo.pl -c 1 data/all_subreads.fasta > wtasm.mak
make -f wtasm.mak

I have messed around with providing all reads and varying degrees of higher quality subsets of reads. It seems like smartdenovo is very sensitive to read quality in ways other tools are not. For example, Miniasm and Abruijn do better with all or most of the reads whereas smartdenovo does better with higher quality subsets of them.

Are there any parameters I should tweak to help smartdenovo exploit all the reads?

best,

John

[ruanjue]

Thanks. In fact, smartdenovo is designed to use high quality subset. smartdenovo uses SW pairwise alignment, if exploit all reads, will slow the assembly. When making consensus sequences, you can rescue all reads for Quiver.

There are many parameters in each step of the pipeline. Try more parameters for wtclp (detecting chimeric reads).