For ancient DNA applications, it would be really helpful to have a modifier in MEGAN where reads could be restricted to customized fragment lengths (say minimum 30, or 35 to help with removing false-positive reads) to test which works best for a particular dataset. I typically use a minimum 24 bp cutoff during bwa mapping prior to blast and megan. But it would be nice if this could then be customized during taxonomic binning rather than needing to go back and re-blast the mapped reads with a new minimum size cut-off.
So, you want to be able to specify min (and max?) length of reads to be considered? Any read whose length is not within range gets put on the “Not assigned” node?
I took another look at this. I don’t really want to add two more features to an already crowded interface… However, what I can do is to replace the “Low complexity” feature by a “min read length” feature so at least short reads can be caught and assigned to the “Not assigned” node. I don’t think that anyone uses the “low complexity” feature and so I will retire it in the future.
In fact, I have already done the programming, just need to test the code a while to avoid breaking stuff…
