how different the rep rates need to be to be considered as such?

[xvazquezc]

Hi there,

we have a very simple system (2 species) and we have been investigating the effects of certain chemical on it. I've run iRep on control and treatment for both species.

Species	Control	Treatment
Sp1	1.54	1.52
Sp2	1.28	1.33

Certainly, Sp1 has a higher replication rate but what I can't say confidently it's if there are actual differences in control vs treatment for any spp. How sensitive is the output to noise? Is there a threshold to say if the replication rates are different? I don't know if this question has been explored but I'd be grateful for any insights you could provide on this.

Thank you in advance,
Xabi

[christophertbrown]

Hi Xabi, These are good questions.

Certainly, Sp1 has a higher replication rate but what I can't say confidently it's if there are actual differences in control vs treatment for any spp.

I agree that Sp1 is replicating faster, but do not think there is a difference between control and treatment for either Sp to claim a change in replication based on this data.

How sensitive is the output to noise?

The biggest contributor to noise is read mapping ‘contamination,’ usually caused by the presence of a related species or strain. It sounds like this should not be a major concern with your system.

Is there a threshold to say if the replication rates are different? I don't know if this question has been explored but I'd be grateful for any insights you could provide on this.

This is an important question that we have unfortunately not yet been able to address. Anecdotally, I would not consider differences < 0.1 to be significant. Chris

…

On Oct 13, 2020, at 11:59 PM, xvazquezc ***@***.***> wrote:  Hi there, we have a very simple system (2 species) and we have been investigating the effects of certain chemical on it. I've run iRep on control and treatment for both species. Species Control Treatment Sp1 1.54 1.52 Sp2 1.28 1.33 Certainly, Sp1 has a higher replication rate but what I can't say confidently it's if there are actual differences in control vs treatment for any spp. How sensitive is the output to noise? Is there a threshold to say if the replication rates are different? I don't know if this question has been explored but I'd be grateful for any insights you could provide on this. Thank you in advance, Xabi — You are receiving this because you are subscribed to this thread. Reply to this email directly, view it on GitHub, or unsubscribe.

[xvazquezc]

Thanks @christophertbrown .

This is interesting esp. bc when we look at the abundances of the two spp. based on read recruitment, there is indeed some clear differences in the composition of the system:

Species	Control	Treatment
Sp1	18.9%	9.1%
Sp2	80.1%	90.9%

Cheers,
Xabi