Hi from Chapel Hill, NC

[wjeck]

Hi everyone,

I'm new to the forum today and I love what I see. I've been in next gen sequencing for about two years. I worked on VCAKE, and it's still alive and kicking, now optimized for 454/Illumina bacterial genome assembly ... pipeline forthcoming at vcake.sourceforge.net hopefully in the next few weeks. I can't claim it as my work, but I might as well pump it up a little.

Now I'm an MD/PhD student looking to apply the new genomics to clinical questions (particularly oncology). I've just spent 2 years in medical school so I am all but completely out of the loop, but I'm looking forward to getting back up to speed.

--Will Jeck
UNC Chapel Hill
School of Medicine

[yoyoq]

hi,
i have some vcake questions.
does vcake consider both strands of the input sequence?

i have short reads(31 long) , in my test case they all blat perfectly
to a known genome and there seems to be decent overlap.
but the vcake is not tying the contigs together, i am getting a
max contig of 100 for a 1000 base region
do you have any suggstions?

thanks
jpd

[wjeck]

The short answer is yes, it does consider both strands.

My recommendation would generally be not to use VCAKE unless you are doing hybrid assemblies of 454 and Illumina data. VCAKE has not been maintained to deal with updates in the technologies, unlike some other algorithms. I would refer to this thread and try algorithms like Velvet, which make use of a more sophisticated graph decomposition method.