search

borenstein-lab / fishtaco

FishTaco (Functional Shifts Taxonomic Contributors) is a metagenomic computational framework that aims to identify the driver taxa of microbiome functional shifts
Other
23 stars 4 forks source link

How many rows should I choose about function (e.g., KO) records in Function abundance file #1

Open PingPi357 opened 7 years ago

PingPi357 commented 7 years ago

Hello, I am using your tool, thanks for your sharing. The programme proceeded with error when the Function abundance file had several (e.g., 1, 2, 3, 5) records (error: .Unexpected error while running MUSiCC on functional data (make sure you have KEGG orthology groups (KOs) as rows): (<type 'exceptions.ValueError'>, ValueError('Cannot have number of folds n_folds=5 greater than the number of samples: 1.',), <traceback object at 0x7f70f5b832d8>) ), but it worked well more than 200 records. I wanna know why and is there any relationship between different records?

omanor commented 7 years ago

It seems like you only have one sample in your file. FishTaco is a tool to compare 2 sets of samples to each other (e.g., cases and controls), so I'm not sure what exactly are you trying to apply it to. Can you please give a little more information?

PingPi357 commented 7 years ago

I have solved the proplems and it is because of the unexpected input, and I still have a question about the 2 sets of samples (cases and controls), if I switch the position of the samples (for example, pretreatment for control 0 and posttreatment for case 1 or pretreatment for control 1 and posttreatment for case 0), I got a completely different results. It is hard to interpret it and what is your idea?I am looking forward to your reply, thanks very much. Yours!

engal commented 6 years ago

Hi, I'm taking over providing support for FishTaco. Could you give an example of what you mean by completely different results?