Title: "Issues with Converting .raw Basis Sets to .BASIS File Using LCModel's makebasis Function


I want to convert .raw basis sets into a .BASIS file. I am facing difficulties in generating the .BASIS file using makebasis.in provided by LCModel package itself. I have verified that the paths in the makebasis.in file exist and are correct.


I am using CentOS Linux release 7.9.2009 x86_64, and the version of LCModel I am using is lcm-32 v6.3. The reason for using lcm-32 is that lcm-64 could not function properly after installation, and after installing lcm-32, it worked fine.

Has anyone encountered such a situation? I would greatly appreciate your assistance.

Hi @Yushan,

The first part of the error message suggests that one of the files mentioned in your makebasis.in could not be found… I guess it might be the output folder

If your makebasis.in, check that the folders mentioned in filbas and filps exist. It looks like you may have specified these as relative paths, but you’re running makebasis from within the bin subfolder so it’s not finding them.

So: try specifying the full path for all these items in your makebasis.in… probably something like: /home/yushan/.lcmodel/test/output/basis.ps

If that doesn’t help, make sure the folder actually exists (mkdir -p ~/.lcmodel/test/output), and also check the other filraw lines (.RAW files your makebasis.in is referring to).

Hope this helps.

Oh. I just noticed this. Maybe double-check the relative paths still, but if that’s all fine perhaps you could share the first couple of “paragraphs” from your makebasis.in file?

If you have a folder of individual .raw files for each metabolite (as generated by FID-A simulations), you can also try the fit_makeBasis function in our Osprey repository to create an Osprey-style basis set, followed by the io_writelcmBASIS function.

Thank you Alex! It’s exactly the reason that caused the problem.

By the way, if the file path is ~/.lcmodel/test/output/basis.ps, it still can’t be found. The command only runs successfully when the file path is like /home/yushan/.lcmodel/test/output/basis.ps.

1 Like

Hi, Georg

Yes, my basis sets are generated by FID-A simulations. I tried the fit_makeBasis function and the io_writelcmBASIS function. It’s convenient.

Here are a few notes about creating MEGA-PRESS basis sets, just for reference.
1.As the last step of FID-A simulation, save both outON and outOFF in one .mat file for each metabolite.
2.Check the Matlab paths. Remove FID-A folders. (Functions such as 'op_gaussianPeak' share the same name in FID-A and Osprey, but have different content. )
3.When using the 'io_writelcmBASIS' function to create ‘diff’ and ‘off’ basis sets, one may make the following modification:
In the line RF=shift_centerFreq(basisSet,i), change it to RF=shift_centerFreq(basisSet,i,subSpec). For the parameter subSpec, input 1 for ‘off’ or 3 for ‘diff’.