Using LoFreq-NQ?

[fedarko]

Hi, and thanks for developing LoFreq!

We've been looking into using the "LoFreq-NQ" approach described in Wilm 2012, which doesn't take Phred quality values into account and instead uses a 12-parameter model based on the 12 possible nucleotide substitutions. (The main motivation is that the dataset I'm working with right now doesn't have reliable Phred quality values.)

However, I can't seem to find an option for using LoFreq-NQ, or otherwise ignoring individual bases' Phred quality information, when running LoFreq (installed version 2.1.3.1 via bioconda). I can see various options for ignoring BAQ values, and I can see options for filtering bases based on their qualities, but I can't see an option to just ignore all positions' base quality values.

I can see that there is code for LoFreq-NQ distributed with LoFreq version 0.1 on SourceForge, but I know LoFreq has been developed a lot since this version's release in 2012 :) Is there a recommended "best practice" way of currently using LoFreq-NQ, or has the functionality in LoFreq-NQ mostly been deprecated / integrated into the "main" version of LoFreq / etc.?

Thank you!

[LaraFuhrmann]

Hi Marcus,

I am in a similar situation like you. Were you able to run lofreq ignoring all positions' base quality values? If so, how did you do it? :)

[fedarko]

Hi Lara,

As a not-ideal solution, I ended up just running a recent version of LoFreq on our data (the exact script we used for this is located here), without explicitly using LoFreq-NQ.

[LaraFuhrmann]

Hi Marcus, thanks a lot for your speedy reply and sharing your solution. I guess, I will also just run LoFreq on my data. Thanks again!