These are chat archives for dereneaton/ipyrad

Mar 2018
Paolo Momigliano
Mar 20 2018 11:17
Hi Isaac and Deren. In understand that in release 0.7.16 you fixed a bug causing lots of rm_duplicates when analysing SE RAD data. I do not quite understand: what exactely are rm_duplicates? Also, we analysed a fairly large dataset (130 samples, normal RAD) using a version of iPyRAD pre-dating release 0.7.16, and indeed a lot of the of loci (85%) were filtered by rm_duplicates according to stats file. If we re-run the pipeline with the latest version (i have tried with 10 individuals only to check), only about 3% of the loci are removed by rm_duplicates. Does that mean that the dataset produced by the previous versions of the pipeline were not reliable? Any help would be very much appreciated. Paolo
Mar 20 2018 12:28
Hi there, it seems I cannot update ipyrad to version 0.7.23. Although I get the information "All requested packages installed", only 0.7.22 will run afterwards. Could it be a bug?
Eaton Lab
Mar 20 2018 14:17
Hi @PaoloMomigliano_twitter, that bug only affects reference mapped SE RAD data, not denovo. If that is your data type then yes, I would recommend re-running it. Before v.0.7.16 we had only optimized the reference-based methods for paired ddRAD data and so this bug caused many loci in single-end RAD data to be excluded erroneously.
Paolo Momigliano
Mar 20 2018 14:35
@eaton-lab Thank you for the prompt reply, and yes the data were indeed reference mapped SE RAD. Does that means that also the "good" loci retained in our previous analyses would be problematic, or simply that lots of good loci would have been discarded?