I have a TMT experiment with several combined samples. In each I have a pooled reference sample to allow IRS normalisation. Whilst individual samples have a specific biological replicate number, pooled reference samples are a mixture of different replicates so it would be wrong to assign the replicate number one to them. However, values of "not applicable" are rejected by the validation tool. For now I will put 1 there but this is potentially misleading and a weakness in the current file scheme.