aggregate_pcrs returns Error message column `type` of table `pcrs` in out must contain only `control` or `sample` values

[lzinger]

error message come from issues when recreating the metabarlist (from check_metabarlist). Occurs with the grouping factor (by default "sample_id" containts NA values. We should either specify it in the help, or find another way to solve this.

[CassandreHeritier]

Hi, we are encountering the same type of problem. With the command data.merged <- aggregate_pcrs(data.samples), we get the following error even though no combinations are identical, for a single sample in four PCR replicates. So we have four rows in the pcrs table that conflict, even though nothing seems to make them identical :

I can send you the example that causes the error in a private chat (which I am not including here for confidentiality reasons). If we test check_metabarlist(data.samples), it returns TRUE, while aggregate_pcrs which calls check_metabarlist() returns the error.

We are in trouble with this problem because it is an important project for a client.
Thanks for your help.

[lionncle]

Hi, at first look, this is not the same error as the one indicated in this issue. Your error message indicates that you have the same combination of plate_no + plate_col + plate_row several times, which is not allowed in our data schema.

Furthermore, it seems to me that the aggregate_pcrs function is not used on the samples dataframe.

I think that my answer will help you solve your problem. If not, can you generate a download link for your data and send it to my email address, which I will provide to you later?

[CassandreHeritier]

Hi,

yes, the name of my object is confusing, but data.samples is indeed a metabar object, with four lists, including a pcrs list on which the error is raised. We do not have any duplicate combinations of plate_no + plate_col + plate_row in the pcrs table, and if I test check_metabarlist(data.samples), I get TRUE, but when I go through aggregate_pcrs(data.samples), I get the error.

We have identified the PCR replicates that are causing trouble, and we use this function on a daily basis without ever encountering any problems with this type of data. The function works correctly on all other PCR replicates; there are only four PCR replicates from the same sample that are causing problems and are unable to aggregate correctly. If I remove any of these four replicates, it works. The format does not appear to be incorrect either.

So I can send you by email the R object if you want to reproduce the error.

Thanks a lot

[lionncle]

Hi, you can send me the R object to the following address : clionnet@chu-grenoble.fr