[westerman]

PBJelly keeps crashing on me. Previous jobs (with potentially poor Illumina assemblies) produced poor results. That said I do have a current job that has been running for three days without crashing so I am hoping for the best. That job is a small variant on jobs that have crashed on me but maybe I got everything right this time. As is typical for bioinformatics program the troubleshooting messages are not very helpful.

I do realize that PBJelly is newer than AHA but one has to do with what actually runs.

Looks like you just removed the AHA option from the web page I posted above. Your first edit of that page.


While I haven't used MaSrRCA in a while it seems to me that it is basically doing an Illumina assembly and then layering -- scaffolding, gap-filling, whatever -- on top of the Illumina assembly. Same category


I point out that while PacBio only assembly is nice it doesn't really work for a 1GB genome unless one has a lot of $$$.

[rhall]

I removed AHA, as I cannnot think of any situation for which I would recommend it. It doesn't work for large genomes (>200Mb), and for smaller genomes if you have enough data to use it you have more than enough to try hybrid assembly, or even low coverage denovo PacBio (>25x).
I would consider scaffolding and resolving ambiguity during assembly to be two different approaches.
Cost for a 1Gb genome is probably not as much as you think, and if you factor in analysis time, it can be competitive. A recent example avian genome, great library, 60 SMRT Cells, from raw data to finished genome ~2 days, >15Mb contig N50. Obviously not everything works out quite so well, but 1Gb genomes are becoming routine.

[rhall]

The Falcon assembler is designed for diploid assembly, for higher ploidy it will depend on the nature of plody inbred, allotetraploid etc.