Technical vs biological replicates

[Danvor]

Hi,

I am trying to run discoRhythm to on transcriptomic data but I am not sure how biological and technical replicates are handled and whether my dataframe set up is wrong.

I made Summarized experiments from a dataframe where the first column is gene names and other columns are counts per time point per biological replicate using discoDFtoSE. I am inluding the the sreenshot of the top of the view(assays(se)[[1]]) that I used to input data.

As you can see I have PrefixTime_UniqID_ReplicateID, and so every column is a different biological replicate per timepoint. But when I check view(discoDesignSummary(se)) or view(discoDesignSummary(colData(discoCheckInput(se)))), I see only in my case 7 biological samples with different, I guess, technical replicates, in the brackets (see screen shot "discoDesignSummary from meta"),

as your vignette states:

The sample collection information present in colData(selectDataSE) can be summarized by the discoDesignSummary function to detail the number of biological and technical replicates available at each collection time. Number of technical replicates is shown in brackets.

So I wanted to ask, what am I doing wrong? Why does discoRhytm not seeing my samples as biological samples? Am I putting the UniqID wrong?

Any help appreciated. Thanks.

DanVor

[matthewcarlucci]

Hello,

From earlier in the vignette (section 4.3):

Biological vs Technical Replicates - Time + Replicate Id are used to identify independent samples collected at the same timepoint (biological replicates). Samples with the same Time and Replicate Id are assumed to be technical replicates originating from a single biological sample.

Or, since you seem to intend to treat all of your data as independent biological samples, you could also omit ReplicateID:

If no Replicate Id is provided, all samples are assumed to be independent biological replicates.

i.e. any of MD11_1NM1, MD11, 11 should work for your case. Hope this addresses your need.

A link to 4.3 in section 6.1.4 probably would have helped you here, so I will aim to add this to the documentation for future users.

Best,
Matt

[Danvor]

Sorry, I misunderstood that part, I thought if you give it a unique ID it would treat them as different.

Removing the replicate IDs worked and DiscoRhythm sees my samples as biological replicates.

Thank you!

Best,

DanVor