GS Junior - new software

[audioPL]

Hi All,

Have you upgreaded GS Junior to the new software package (v 2.7)? I am happy with the results on the old version so I don't want to 'make it better' until I will be sure that it works fine. How it change your results? Is it worth to upgrade?

I look forward to hearing your opinion,

Bests,
m

[audioPL]

So nobody has upgraded to 2.7?

Bests,
Mirek

[rexxi]

Has been released? I'm using just 2.6, if I install it, I'll tell you the deiferences

[yaximik]

I have some experience with 2.7. It has number of new features yet it has some problems as well, primarily in the data analysis part. Data collection and processing works fine, but I encountered some bugs with gsMapper. First, while installation on the attendant PC (CentOS5) runs smoothly, there is some problem with making desktop links on the offrig (RHEL55) for the entire data analysis module. Before I figured out how to create working links in a panel, apps had to be launched from the command line. Recently I experienced abortive runs with gsMapper failing to produce output files due to internal error after essentially completing the job. Software engineers ata 454 have not yet responded with the solution. If I continue to experience this, I rather go back to 2.5p1, which worked just fine.

[audioPL]

Thx for your comment yaximik. Could you tell me please more about general improvements on picture processing and its impact on results? Do you see an increase in the number of hq reads and total bases? As I don't need GSmapper, would you recommend me to uprgade to 2.7?

bests,m.

[yaximik]

This obviously depends more on quality of samples and reagent amounts than on software. No drastic improvements and it was not promised. I did get some excellent runs exceeding benchmark capacity of the instrument, but I also experienced very bad runs as well. With good libraries, read length distribution was closer to maximum possible for GS Jr kits, with sharp drop between 500-550 nt. In overall, i got upgrade for free since I was filed testing it, but if I would have to pay it did not feel enciting. So if it is a free upgrade then why not? The analysis package is a different issue.

[Palecomic]

With 16S we've found that you get different communities of bugs (alterations in relative abundance of each organism) depending on whether you use amplicon or shotgun processing and depending on the software version. Caused us a lot of fun the last time we upgraded. The important thing to do is to make sure that within the same project, everything you want to compare is processed identically, including software version of the system software, all the way through.

I tend to shotgun process as I'd rather retain more reads myself and remove the bad ones using something like QIIIME's denoiser or Chris Quince's amplicon noise which also chimera-checks, than hard cut-offs of the amplicon pipeline.

Have tried to time this update between projects for most people, and everyone else will have to reprocess their data!