Blowups with de novo peptides

GROMACS version: 2021
GROMACS modification: Yes/No - No

I’m doing solution simulations on de novo peptides (under 15 amino acids) I’ve made in Avogadro. The molecules are pushed through the CHARMM-GUI. Parameters are all the same as in this job 1695241522 (this is the job ID) that you can use by going into the job finder on their page. But the most important parameter is that I’m using the CHARMM36m force field which should be optimal for peptides and proteins.
I’m using 5 mil steps and everything else in the configuration files is as the standard CHARMM-GUI input file looks like.

The issue is that all of my molecules have enormous distortions from step 1 when I open them in VMD. The program blows up and I don’t know whats causing it.
I don’t know how to even troubleshoot it.

Im suspecting its about the de novo molecules not being minimized properly but I don’t know how to test that hypothesis. The Avogadro program has a minimization option which I used and then additionally minimized it in GROMACS, but the issue remains.

Please help.