Do people still use PICRUST?

General Discussion
1

Hi all,

I am wondering if people still use PICRUST? I've heard it has been shunned as a predictive tool. I am wondering whether to include it in a paper I have.
If not are there newer functional predictive tools available for 16S data?
Thanks,
Sam

1 Like
2

Hi @Sam_Degregori,

I think there are some people who still use PICRUSt, but it has caveats. (I think they're well described in the PICRUSt II preprint.

One of my questions would be how functional inference will add information. I had a recent project where all my differences were between closely related organisms. In that case, I didn't think functional inference would add much because I was working below the threshold of what could be done with PICRUSt.
You might also want to think about whether or not you're working in a well defined enviroment.

Best,
Justine

2 Likes
3

Appreciate the response. I will check out the preprint.

My experiment has minor yet significant changes in a handful of microbial taxa (16S data). There is a good spread of phylogenetic variance so I don't think I have that same issue.
My only issue is that the pathways spit out by PICRUST are very specific and I am having trouble making broader generalizations about the functional profiles of the communities.

1 Like
4

I have had that exact same issue!

It would be great to have a tool to 'roll up' groups of functions into metabolic pathways. Has anyone found an elegant way to do this?

Is this even biologically feasible or is this one of those complicated problems? :thinking:

Colin

1 Like
5

Old PICRUSt1 did that.
There is a mapping between the levels in the group. Someone would have to build that database and then write a script, but in theory, it's doable.

I think I did hand annotation a while ago based on the ID numbers, and that gave some structure but I didnt end up pursuing it because my interesting things were too closely phylogenetically for PICRUSt to make sense.

Best,
Justine

Best,
Justine

6

just want to toss this paper in the mix... relating to @Sam_Degregori's original question about the predictive accuracy of picrust and similar tools:

2 Likes