I noticed that when using a PEP as input and updating that same PEP with a downstream result using pipestat, that sometimes results will not be reported back or, more alarmingly, existing data will be nulled.
using this pep:
https://pephub.databio.org/donaldcampbelljr/mouse_seqcol_digests?tag=default
and this pipeline: https://github.com/refgenie/refgenomes.databio.org/blob/donald_PEP_additions/scripts/seq_col_comparison/pipeline/seq_col_jsons.py
This happened to a couple of samples and it happened to me twice today on two different PEPs. I'm using looper to run the samples in parallel on the HPC.