[lcollado]

Thanks! cor.plots and confidence.ellipse look interesting as well

[strob]

Hello,

this tool looks really handy IF you have your reference genome sequenced. For people working on plants, this is most often not the case. So I wonder why you use velvet, which is a de novo assembly tool, and not one of the reference based tools.

Do you have a solution for validating velvet results for people doing real de novo assembly?

Many thanks,
Steven

[Zigster]

i am doing de novo txome assembly of plants

basically I just use the closest related species to see if something unexpected is happening with the parameters
most of this short script is concerned with N50 and assembly size

[strob]

I would really like to test your tool, but it seems to use BLAT, which is for us as a company not freely available. Do you think you could incorporate another (complete free) alignment tool?

[Zigster]

Yes I will make the alignment part optional or BLAST-driven.

[bioinfosm]

Thanks Zigster. If you could add a very small data set to right-away test the scripts on a particular person's setup, it would really help!

[bioinfosm]

Here is the plot that I see.. thanks for the handy tool Zigster
comments or observations from the plots?

[Zigster]

strange - it looks like the longer kmers are just hurting you. typically they will produce longer N50's and smaller assemblies.

That is unless they are getting close the length of your reads. How long are your reads?

[bioinfosm]

This is old data with 36bp reads

[Zigster]

wow talk about attrition

might be interesting to see some smaller kmers - 15 and 17

[Zigster]

The Standardized Velvet Assembly Report has been updated to accept BLAST m8 output or no reference genome at all.