These are chat archives for dereneaton/ipyrad

6th
Mar 2017
Emily Warschefsky
@ewarschefsky_twitter
Mar 06 2017 14:57
@isaacovercast - The libraries are ddrad and were prepped exactly the same way (same REs). I reran the PE as SE from steps 1-5, and merged with the true SE lanes before step 6. Step 6-7 ran, but the stats file says that, although many clusters were found within the PEasSE individuals, there are no loci in the assemby. Below I show a sample of what the final stats file looks like - I added a column to note which samples were in which lane. The really odd thing is that, with min_samples_locus set to 9, I should at least see individuals from the PEasSE lane clustering within themselves (since there were 88 of them total, and when run alone as SE or PE they do work), even if the clusters don't align with anything from the other SE lanes. So it seems like it's something to do with ipyrad not being able to find the reads_consens files..
error_est     reads_consens        loci_in_assembly    Lane
0.001776    73562                        NaN             1(PEasSE)
0.001664    56891                        NaN              1(PEasSE)
0.001226    53707                        NaN              1(PEasSE)
0.001817    49250                        NaN               1(PEasSE)
0.001275    49024                        NaN                1(PEasSE)
0.002338    48266                        12135          3(SE)
0.00235    47640                        19174               2(SE)
0.002209    46289                        20055               3(SE)
letimm
@letimm
Mar 06 2017 17:03
@isaacovercast @dereneaton I am having trouble with step 7 (ipyrad version 0.6.4). I am requesting all output file types, but it errors out after writing a few of them (.hdf5, .loci, and .vcf). The error reads:
index -1 is out of bounds for axis 0 with size 0
I have not had this problem before; about a month ago I wrote out all outfiles for this dataset without a problem. Is there a reason ipyrad wouldn't write out the other files (.str, .phy, .nex)?
Deren Eaton
@dereneaton
Mar 06 2017 17:29
@letimm that is a bug that we've fixed. Run conda upgrade -c ipyrad ipyrad
and then rerun step 7
letimm
@letimm
Mar 06 2017 17:41
@dereneaton Thank you!
Jay Dimond
@jldimond
Mar 06 2017 20:57
Hello @dereneaton and @isaacovercast . I was wondering if you could point me to some information on the alignment criteria used by ipyrad during the reference mapping step. I could not find any specifics in the online documentation - did I miss something? Thanks in advance!
letimm
@letimm
Mar 06 2017 21:56
@dereneaton @isaacovercast I have run into a different problem. On step 3, dereplication occurs, but clustering reaches 62% and goes no further. I am only having this problem with one of the datasets I'm working with, the others seem to be running fine. The ipyrad_log file is empty and there is no error message. Any ideas on why it is getting "stuck"?
Deren Eaton
@dereneaton
Mar 06 2017 22:36
hey @letimm, the step3 clustering progress bar is showing the number of samples that have finished clustering. If it is at 62% that probably means something like 62/100 samples have finished, while the others are still running, or waiting to start running. You can check if it is still running or has truly stuck by either using a tool like top to see which programs are running, or by checking whether the file sizes in the [name]_clust_[0.xx]/ directory are increasing. If the number of reads in some samples is much greater than others then that can cause the progress to increase at an uneven rate.
Alternatively, if it did freeze and nothing appears to be running then we can try to figure out why.