These are chat archives for dereneaton/ipyrad

Apr 2017
Edgardo M. Ortiz
Apr 17 2017 16:49
Hi @dereneaton and @isaacovercast, I have a question regarding parameter maxalleles [18]: I have an old analysis in pyrad where I assumed maxalleles=2 for the seven steps, and when using the same settings in ipyrad I couldn't recover a comparable number of loci (ipyrad was getting rid of many that made it all the way through pyrad). Then, reading the docs I realize that now maxalleles also affects step 7, so I relaxed the parameter to 4 to re-run step 7 alone and just then I could recover a similar set of loci between both pyrad and ipyrad. If I already used maxalleles=2 for steps 4 and 5 in ipyrad, how is this parameter affecting step 7? I would assume that by that point (s7) the consensus sequences that were clustered all passed the filter of having at most two alleles, am I overlooking anything?
Matt Johnson
Apr 17 2017 19:55

Hello! I am running iPyrad for the first time on our own data, after successfully navigating the tutorial. I have single-end data for 32 individuals in four populations. Unfortunately, the data came back to me already demultiplexed, but I ran it through Step 1 and Step 2 without any errors. Step 3 took about 3 days on our cluster and returned a number of errors, I believe having to do with the clustering:

2017-04-16 14:59:16,173 pid=6963 [] ERROR sample [Poland_Ohar_MSR-01_GTGCACCA_1] failed in step [build_clusters]; error: ValueError(need more than 1 value to unpack) 2017-04-16 14:59:16,174 pid=6963 [] ERROR sample [Poland_Ohar_MSR-09_TGGC_1] failed in step [build_clusters]; error: ValueError(need more than 2 values to unpack)

There were nine individuals with this warning, each of which had many thousands of reads pass the filtering step. I am trying to figure out if there is something wrong with our data or our iPyrad installation. As best as I can tell, the samples which have errors do not have completed .clustS.gz files, and instead have several intermediate files, such as utemp.sort, .utemp, .htemp remaining. A quick view of these temporary files seems to show many clusters, although I'm not sure what I should be looking for. What would be a good way to figure out if it is a problem with my data or with iPyrad?

Anna Mason
Apr 17 2017 23:31
@isaacovercast Ok thanks Isaac, that does make sense!