velvet assembler and paired reads: -shortpaired vs -longpaired?

Question

[jeremy.cox.2]

Hello all,

I have been using oases pipeline (and therefore, using velvet as a base program) for single read data.

Now I have paired end data (150bp), so I am trying to decide what settings to run and use.

I have gotten stuck because velvet allows multiple types of data, but it is not clear how to declare my data type.

Specifically, I have the choice between "-shortPaired" and "-longPaired".

I have been reading documentation, and it is all silent on the subject. (or I misunderstood the documentation.)

The only reference I found is in the program --help on command line:

velveth Assem 43 -short -fastq unmapped.fna -longPaired -fasta SangerReads.fasta

Here, long reads are tied to the idea of Sanger-style reads. Sanger implies to me hundreds of basepairs, but I have seen the cut-off in software and literature sometimes be 100, 125, or 150.

Also, the documentation implies that long reads can be used to check and to break missamblies during velvetg step. If this is the case, I think it would be only appropriate to include "long" reads when I have two sequencing methods, one of which contains long sequences (contrasted to the short).

So what difference does this option make? Which setting should I use?

Answer 1

[pld]

It says in the documentation under 5.6 "What's long and what's short?":

"Velvet was pretty much designed with micro-reads (e.g. Illumina) as short, and short to long reads (e.g. 454 and capillary) as long."

The documentation appears to suggest that illumina data be flagged as short reads.