These are chat archives for dereneaton/ipyrad

21st
Aug 2016
Katia Capel
@katiacapel
Aug 21 2016 00:47

Hi @isaacovercast, thanks for your answer! I thought that the total prefiltered loci was showing only the shared loci, now it makes sense.
About the other question, I did change the Min # of samples per locus to 3 (I also tried with 2) and I still don't get any output loci. I get this message:
Step 7: Filter and write output files for 3 Samples
[####################] 100% filtering loci | 0:00:13
[####################] 100% building loci/stats | 0:00:01

max() arg is an empty sequence

Cycadales
@Cycadales
Aug 21 2016 00:59
Hi guys quick question has anyone used fineRADstructure with the outputs from ipyrad? if so which output file would you use or modify to use? I am trying to find a program like Structure to deal with rad data.
Isaac Overcast
@isaacovercast
Aug 21 2016 16:29
@katiacapel - You might also have to reduce the value of the trim_overhang parameter as well. Will you send me your params file so i can take a look? iovercast@gc.cuny.edu
Isaac Overcast
@isaacovercast
Aug 21 2016 16:34
@Cycadales Hey James, i looked at fineRADstructure and the input file format they want is not very similar to anything we write out. You might try faststructure (reads .str format) or sNMF (which reads .geno i think)
Cycadales
@Cycadales
Aug 21 2016 16:38
@isaacovercast Yea I tried the .str formats without much luck...so I will the the geno and see what it gives.
Isaac Overcast
@isaacovercast
Aug 21 2016 16:39
Yeah, try sNMF it's really fast.
Cycadales
@Cycadales
Aug 21 2016 16:40
Have you had much luck using faststrucutre?
Isaac Overcast
@isaacovercast
Aug 21 2016 16:41
I haven't ever used it personally. I know people who've used it.
Komlan Avia
@coolkom
Aug 21 2016 22:10
Hi all. I have been analyzing a ddrad dataset with 320 samples across several species. It took more than 2 days to finish but then I got this error: Unable to create dataset (Chunk size must be < 4gb). This likely happened at the step 7 since I can see results of all the previous steps. Can someone help me understand this and would it be possible to create the final result files starting from where the analysis stopped without restarting? Thanks.
Isaac Overcast
@isaacovercast
Aug 21 2016 22:41
Hi @coolkom I'll check it out. Assuming this problem is only related to step 7 we should definitely be able to pick up where the analysis left off. Can you re-run ipyrad just for step 7 and include the -d flag. This will generate a file called ipyrad_log.txt. If you can email that to me that'd be great: iovercast@gc.cuny.edu
Komlan Avia
@coolkom
Aug 21 2016 22:43
Thanks @isaacovercast. I'll do that and let you know.
Isaac Overcast
@isaacovercast
Aug 21 2016 22:43
@coolkom Also, what OS are you running on? Is it a cluster or a desktop?
Komlan Avia
@coolkom
Aug 21 2016 22:43
It's a cluster
Isaac Overcast
@isaacovercast
Aug 21 2016 22:49
@coolkom Can you also send me the output for ipyrad -p <yourparamsfile> -r, i'm pretty sure step 6 actually probably failed.
Komlan Avia
@coolkom
Aug 21 2016 22:50
ok
Can I send that one here ?
Isaac Overcast
@isaacovercast
Aug 21 2016 22:55
Sure. you can PM it to me
Komlan Avia
@coolkom
Aug 21 2016 23:00
Done
Isaac Overcast
@isaacovercast
Aug 21 2016 23:02
Cool I think i know the problem. What does your data look like? PE ddrad? GBS?
Komlan Avia
@coolkom
Aug 21 2016 23:03
normally I have paired-ddrad
Isaac Overcast
@isaacovercast
Aug 21 2016 23:03
Got the -r output. I see all the samples are still in state 5, so step 6 definitely is where it's failing.
Komlan Avia
@coolkom
Aug 21 2016 23:04
but I concatenated R1, R2 and singletons I got after Stacks first steps into a single file per sample
OK you're right then
Isaac Overcast
@isaacovercast
Aug 21 2016 23:05
Got it. I definitely know the problem. It could take a little time to fix. I'll get back to you later tonight or tomorrow.
Komlan Avia
@coolkom
Aug 21 2016 23:08
Great. Thanks a lot.