NGS Analysis pipeline comparison

[MenzZana]

I am looking at what software to use for NGS data.
At the moment Galaxy seems the software of choice by most people.
Although Galaxy has some very nice tools, its workflow sharing and API are far from perfect.
Also if you want to make batch analysis of NGS data, a pipeline software would be preferred.

On this matter I am looking closely at mainly 2 softwares that can be integrated into Galaxy, namely KNIME and Taverna.

At a glance, KNIME seems the more mature of the products. It has bioinformatics/NGS algorithms that can be added. Suitable for a HPC environment, even if it comes with a prizetag, and has a whole range of other tools for other types of analysis.

Taverna is solely focused on bioinformatics.

You have also a platform called YABI, which can be used from a webinterface and seems like something in between Galaxy and pipeline software, namely you can construct a pipeline but as I can see not fork the analysis so its quite restricted.


Has anyone any in depth experience with these software and can give me their opinion which does suit best for large scale NGS?

[GenoMax]

I do not think there is a simple answer for the question you are asking.

The answer is going to depend on the scale you are looking at in terms of samples (tens, hundreds or more), locally available expertise for implementing these programs (e.g. galaxy is not the easiest to implement on a local cluster, as you seem to have discovered already), a need to make this available through a webinterface for end-user access.

SeqWare is one project you may want to look at. http://seqware.github.com/ It is not for the faint of heart and will likely require time investment on your part.

[MenzZana]

Thanks! I will take a closer look to that one as well