Hi, The abundance counts for the M5NR table (your question two) have been corrected and are now consistent with the barchart from the workbench proteins, thanks for informing us about this. Regards, Mark On Jun 24, 2011, at 11:19 AM, Mark D'Souza wrote:

Hi,

...

1) I cannot download the data from workbench. It's similar to the problem I asked last time, but not exactly the same. I selected all the annotated data to workbench, and then I tried to download them as the fastafile. However, it cannot be done. The webpage will go back to the main page of MG-RAST or to a page with error message. I doubt if it's related to the size of the file again.

Unfortunately this is related to the earlier problem you had, we are looking into it.

...

2) I found the abundance of certain groups are not identical under different functions. What I did was to select the archaea hits (1372 hits and the abundance is 4576) to workbench, and then, I used these hits to make barchart under functional classification. However, the abundance doesn't match these numbers. For example, The first classification on barchart is Cluster-based subsystems, which contains 5964 sequences. That is already more than the total abundance, 4576. And it is less than the barchart from the whole dataset (39830), so I don't know where these numbers were from.

We are investigating, it looks like the abundance count is not being displayed correctly then the M5NR is selected.

...

3) This is just a suggestion. Can you design a filter which is related to the abundance? For example, if the abundance is lower than a number, we can ignore them while annotating.

I believe we have this implemented in the tables, you can specify a lower (or upper) cut-off for the abundance counts displayed.

Sorry for the delay in responding.

Regards, Mark

On Jun 13, 2011, at 10:23 PM, Chih-Ying Lay wrote:

...

Dear Sir/ Madame:

Sorry, it's me again. I still found several problems while using MG-RAST, so I would like to let you know if they can be resolved.

My working metagenome is 4449450.3, and all the problems happened while playing with this dataset.

Here are the problems that I had:

1) I cannot download the data from workbench. It's similar to the problem I asked last time, but not exactly the same. I selected all the annotated data to workbench, and then I tried to download them as the fastafile. However, it cannot be done. The webpage will go back to the main page of MG-RAST or to a page with error message. I doubt if it's related to the size of the file again.

2) I found the abundance of certain groups are not identical under different functions. What I did was to select the archaea hits (1372 hits and the abundance is 4576) to workbench, and then, I used these hits to make barchart under functional classification. However, the abundance doesn't match these numbers. For example, The first classification on barchart is Cluster-based subsystems, which contains 5964 sequences. That is already more than the total abundance, 4576. And it is less than the barchart from the whole dataset (39830), so I don't know where these numbers were from.

3) This is just a suggestion. Can you design a filter which is related to the abundance? For example, if the abundance is lower than a number, we can ignore them while annotating.

So....theses are my questions and problems for now. I really appreciate that you provide us such a good site!

Thank you!

Best Wishes,

Chih-Ying Lay