Celera Assembler 8.2 in SMRT Analysis 2.3

[rboissy]

SMRT Analysis 2.3 uses Celera Assembler 8.1, but a newer version of CA (8.2) is available. In the release notes, it states that CA 8.2 is backward-compatible with 8.1.

http://wgs-assembler.sourceforge.net/wiki/index.php/Version_8.2_Release_Notes

Is replacing CA 8.1 with 8.2 in SMRT Analysis 2.3: (a) not worth the trouble; (b) not advisable; (c) reasonably straightforward? If (c), any advice as to the simplest way to do the replacement (or warning about mistakes to avoid)?

[GenoMax]

I would vote against it (unless Dr. Hall from PacBio tells us otherwise).

[rhall]

(a) not worth the trouble.

As far as I am aware the major changes for 8.2 are the implementation of MHAP for raw read overlapping during correction and pbdagcon as the consensus algorithm in unitig consensus. None of this would not be used by SMRT Analysis, HGAP carries out correction independently of CA using blasr as the overlapper, and pbdagcon is part of SMRT Analysis anyway and default HGAP uses pbdagcon from pacbio code to generate consensus from Celera unitigs.

I would recommend installing 8.2 outside of SMRT Analysis and giving it a try independently. The MHAP correction is very quick and assembly results are comparable with HGAP.

[colindaven]

I would second rhall's comments on MHAP. The correction is quick and results are good for one model organism (A. thaliana) using public SRA pacbio data. I haven't been able to compare to HGAP though yet because of incompatible data formats from the SRA for HGAP.

MHAP is not too tricky to install either.