issue running Velvet optimizer

[akshaya.ramesh]

Hello,
I have been using Velvet to perform a denovo assembly and recently came across the Velvet Optimizer program that would help optimizing the input parameters for Velvet to perform a good assembly. I seem to be running into the same error while running Velvet Optimizer for the past few days.

I use the following command line to run the perl script:

VelvetOptimiser.pl -s 21 -e 91 -m 0.001 -o -exp_cov 0.000001 -f -bam /windows/D/genomes/macaca/paired_merged_Ig_TCR.bam -d /windows/D/genomes/macaca/Velvet_optimized

and the following error shows up:

Minimum specified coverage cutoff is higher than the expected coverage. Please choose a minimum coverage cutoff smaller than 0 and re-run.

I have tried several modifications (change m, the expected coverage, change the default values in the actual script) and different data sets, but I have not been able to proceed further/get rid of the error.

My background in linux/computer science is a month old, so my question might seem very basic and I apologize in advance. Any help would be appreciated Thanks in advance for your time!

[mastal]

issue running Velvet optimizer

Hi,

-m specifies the minimum coverage cutoff, the default value is 0.

As the error message says, you have chosen a coverage cutoff (-m 0.001), which is higher than the expected coverage (-exp_cov 0.000001).

When you run velvetg, if you select -exp_cov auto, -cov_cutoff auto, velvet sets the -cov_cutoff to half the value of -exp_cov.

What are you trying to assemble, and why are you using such a low value for -exp_cov?

best wishes,
maria

[akshaya.ramesh]

Hello,

Thank you so much for getting back to me.

I understand that -m specifies the minimum coverage cutoff. I am trying to optimize and determine the best parameters that would help me perform denovo assembly of the macaque genome where the expected coverage cutoff is 50. However, when I actually run Velvet the coverage of the contigs generated are not anywhere near 50. Also, when I try running Velvet optimizer with no specification of -exp_cov; I still get the same error message.

These are the commands that I use:
VelvetOptimiser.pl -s 21 -e 91 -m 0.001 -o -exp_cov auto -f -bam /windows/D/genomes/macaca/paired_merged_Ig_TCR.bam -d /windows/D/genomes/macaca/Velvet_optimized

VelvetOptimiser.pl -s 21 -e 91 -m 0.001 -f -bam /windows/D/genomes/macaca/paired_merged_Ig_TCR.bam -d /windows/D/genomes/macaca/Velvet_optimized

Initially I did not specify a minimum coverage cutoff/expected coverage. I reduced the mim covergae cutoff to 0.001 just to see if that would get rid of the error message. It was silly of me to reduce the exp coverage below the min coverage cutoff though. I am not sure about what I should change/specify when I run the optimizer to get it to analyze the data.

[mastal]

issue running Velvet optimizer

I would try running VelvetOptimiser with your calculated value of -exp_cov, and leave -m at the default value of 0.

see the formula in the velvet manual for calculating expected coverage using the kmer lengths.

[akshaya.ramesh]

Thank you for the help. I wrote to Dr. Torsten Seemann and he told me that I was running into issues because my -m and -o options were not in quotes as well. I changed the exp coverage, and now that my -m and -o options are in quotes the script is now running!