These are chat archives for dereneaton/ipyrad

12th
Mar 2017
elviscat
@elviscat
Mar 12 2017 13:10
@dereneaton I have a bunch of trimmed data without barcode informationm, is it possible to run ipyrad from STEP 2?
Deren Eaton
@dereneaton
Mar 12 2017 17:23
@elviscat Yes, just enter your data in the sorted_fastq_path, and set filter_adapters to 0. Your Samples will be named according to their filenames. Then run steps 1 and 2. The first step will simply load in the data files, and step 2 will not trim your data any further, but will check it for filter_min_trim_len and for max_low_qual_bases, and format it properly to run step 3 next.
mtcthome
@mtcthome
Mar 12 2017 19:21
@isaacovercast thanks for the reply. I'm working with phylogeographic data at a very shallow timescale. If I use one SNP per locus I end up loosing a lot of information whereas if I use the phased loci I get way more resolution just by having a larger number of alleles. I'll have to stick with the good old pyrad for now, which is also great but takes a while to run because my datasets are large and I have no access to clusters in my university (I'm using my desktop). Any chance I can run some steps in ipyrad and use pyrad for latter steps? Will the outputs of the two pipelines match? Thanks again!
Isaac Overcast
@isaacovercast
Mar 12 2017 21:51
@mtcthome Internally within ipyrad we use a totally different mechanism for handling data, so there's not a way to switch back and forth. I guess it's theoretically possible to run ipyrad through step three and then run pyrad the rest of the way but i would not recommend it.. What kind of analysis are you trying to do?
Deren Eaton
@dereneaton
Mar 12 2017 23:11
@isaacovercast @mtcthome he's looking to get some alleles from RAD data, which we should be able to do. We just need to write a fix to retain lower case bases through the step 6 muscle alignment, which otherwise clobbers them right now. Then just make sure the lowercase bases don't mess up anything else in step7, which I don't think they should..., and then write the alleles similar to the loci file, and make sure .loci is still all uppercase. I'm not gonna have time to do it this week, but it shouldn't take long when we have a chance to do it.