Sourcetracker for sample groups

[alv...@brasaodopampa.com.br]

I have samples of meconium, mother_swab and mother_faeces. I was wondering if I could use SourceTracker to group these samples according to individuals, i.e.
swab 1, meconium 1, faeces 1
swab 2, meconium 2, faeces 2

In a way to consider swab 1 and faeces 1 as sources for meconium 1 and so on. The only way I can think of is doing the small groups individually...that would demand a lot of work.

[Jai Ram Rideout]

Hello,

We have contacted the SourceTracker developers and someone will follow up here shortly.

Best,

Jai

[Will Van Treuren]

Hi,

Unfortunately, there is no option or flag to allow this kind of analysis. I'd suggest using a little bash scripting (if you are using SourceTracker1) or some python scripting if you are using SourceTracker2. 

With SourceTracker1, you'd want to use `split_otu_table.py` based on some mapping file category which gave you separation you wanted, and then you could do a simple bash for loop like

```bash

for table in /path/to/first/otu_table.biom /path/to/second/otu_table.biom

do

    sourcetracker -i $table -m /path/to/mapping/file.txt -o $table+.output

done

```.

With SourceTracker2, you could do the something very similar but it'd be easiest to do the scripting in python.

Hope that helps,

Will