These are chat archives for dereneaton/ipyrad

28th
Feb 2018
George Pacheco
@Pacheco_Evo_twitter
Feb 28 2018 13:46

Dear ipyRAD Community,

I believe I have a special type of data. I have been merging GBS and Re-Seq data and I have everything already mapped to a REF genome, hence BAM files. Thus, could I ask if it would be possible somehow to skip and demultiplexing, clustering and mapping steps and only use ipyRAD to output the found haplotypes? Would that be a conceivable feature?

Many thanks in advance, George.

Isaac Overcast
@isaacovercast
Feb 28 2018 14:16
@Pacheco_Evo_twitter Hi George, mm, this is not something I'd thought about before. There's not really a way to jump into the process mid-stream, or at least there's not a way that doesn't involve some hacking. You could conceivably convert to a sam file, and then run ref_build_and_muscle_chunk by hand for each sample, but this would be an advanced end users only kind of scenario, not for the feint of heart.
George Pacheco
@Pacheco_Evo_twitter
Feb 28 2018 14:28
Hello @isaacovercast Thanks very much for your swift reply! I acknowledge the uniqueness of my query. Alright then, I will try to do what you suggest.
Your best, George.
JStarrett
@JStarrett
Feb 28 2018 18:42
Hello. Just out of curiosity, are the default values for parameters like max Ns, max SNPs, max indels, etc. most appropriate for a specific read/locus length? I am using paired end 150bp, so am wondering if relaxing some of these a little would be acceptable? Thanks!
Gissella Pineda
@gissella_pineda_twitter
Feb 28 2018 20:55
Hello. I want to know if is normal that ipyrad process be slow at step 5 (Conses calling). I mean, the process 3 days ago is at 77% at conses calling of step 5. Mi data set is around of 4 Gb and I am using a cluster of 8 cores. The process doesn't show errors, just is running but is in the same percent. Thanks
Deren Eaton
@dereneaton
Feb 28 2018 21:02
@gissella_pineda_twitter sounds unusual, perhaps there was some kind of silent error. If you interrupt it and then restart step 5, without using the -f flag, it will skip the samples that have already finished and continue running just the unfinished samples.