GAJOE & FFMAKER Errors When Using Both NMATOR and External PDB Pools in EOM

[ddawood27]

Hello,

I’m having trouble running GAJOE through the EOM pipeline using both externally generated and NMATOR-generated PDB pools. I’m using ATSAS 4.1.0-1 on Linux via both the GUI and command line. My experimental data file is correctly formatted with three columns.

First, I attempted to use an external pool of properly formatted PDB files through EOM’s “Insert Existing Pool” feature. When I ran FFMAKER, it failed to generate a size.csv, and GAJOE returned:
error: unable to read intensities file: intensities.csv
error: please check the number of pools

To avoid possible formatting issues, I then used NMATOR to generate a pool of 342 conformers. I ran FFMAKER via:
ffmaker -o intensities.csv --pool size.csv Ma_q_Iq_Ierr.dat mod1_ch/
This created both intensities.csv and size.csv, but GAJOE gave the same error. On inspection, I found that intensities.csv contains only one I(q) curve, despite 342 .pdb files in the pool. It seems FFMAKER is not processing the entire pool.

I’ve ensured there are no spaces in paths, confirmed all files are accessible, and tried both GUI and CLI. Any help is greatly appreciated!

Thanks!

[franke-biosaxs]

I ran the following and it works just fin (on macOS, all path components removed):

nmator --pool 6lyz.pdb
ffmaker --pool=pool.csv  -o intensities.csv nma_odir001/mod1_chA/*.pdb
gajoe -i intensities.csv -s pool.csv lyzexp.dat

Is it possible that "mod1_ch/" does not contain the model files directly? Note the path I used. If that's not it, we will have to look more closely.

[franke-biosaxs]

Looking at this again:

ffmaker -o intensities.csv --pool size.csv Ma_q_Iq_Ierr.dat mod1_ch/

(a) do not include your experimental data in the ffmaker call (only models)
(b) mod1_ch/ is a directory, not model files; try mod1_ch/*.pdb