Hi Maciej,
I've never used MCPB but we're seeing some issues in ParmEd arising from
last month's numpy 2.3 release. Downgrading to a 2.2.X version is helping
people get around some of those errors and may also help here.
https://github.com/ParmEd/ParmEd/issues/1406
Cheers,
Jeff
On Fri, Aug 1, 2025 at 8:24 AM maciej.spiegel via AMBER <amber.ambermd.org>
wrote:
> Hello Everyone,
>
> I am following the MCPB.py tutorial and encountered an error at the
> step:
> "MCPB.py -i MCPB.in -s 2"
>
> The program starts as expected and parses the input successfully, but
> then fails during the angle parameter generation using the Seminario
> method, with the following error:
>
> """
> ******************************************************************
> * Welcome to use the MCPB.py program *
> * Version 2024 *
> * Author: Pengfei Li *
> * Merz Research Group *
> * Michigan State University *
> * AND *
> * Hammes-Schiffer Research Group *
> * Yale University *
> * AND *
> * Li Research Group *
> * Loyola University Chicago *
> ******************************************************************
> The input file you are using is : MCPB.in
> >>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>
> The following is the input variable you have:
> The variable ion_ids is : [1]
> The variable ion_info is : []
> The variable ion_mol2files is : ['PT.mol2']
> The variable original_pdb is : CIX.pdb
> The variable add_bonded_pairs is : []
> The variable add_redcrd is : 0
> The variable additional_resids is : []
> The variable anglefc_avg is : 0
> The variable bondfc_avg is : 0
> The variable chgfix_resids is : []
> The variable cut_off is : 2.8
> The variable force_field is : ff19SB
> The variable frcmod_files is : ['ISX.frcmod']
> The variable gaff is : 1
> The variable group_name is : CIX
> The variable ion_paraset is : 12_6 (Only for the ions using the
> nonbonded model).
> The variable large_opt is : 1
> The variable lgmodel_chg is : -99
> The variable lgmodel_spin is : -99
> -99 means program will assign a charge automatically.
> The variable naa_mol2files is : ['ISX.mol2', 'WAT.mol2', 'CL.mol2']
> The variable scale_factor is : 1.0
> ATTENTION: This is the scale factor of frequency. The
> force constants will be scaled by multiplying the square
> of scale_factor.
> The variable smmodel_chg is : -99
> The variable smmodel_spin is : -99
> -99 means program will assign a charge automatically.
> The variable software_version is : gau
> The variable sqm_opt is : 0
> The variable water_model is : OPC
> The variable xstru is : 0
> ******************************************************************
> * *
> *===================Generate the Initial frcmod file=============*
> * *
> ******************************************************************
> Atoms which has changed the atom types: [1, 2, 18, 34, 36]
> 1-PT.1-PT : PT --> M1
> 2-ISX.2-N1 : n2 --> Y1
> 3-ISX.18-N1 : n2 --> Y2
> 4-CL.34-CL : Cl --> Y3
> 5-WAT.36-O : OW --> Y4
> ******************************************************************
> * *
> *=========Using the Seminario method to solve the problem========*
> * *
> ******************************************************************
> =======================Generate the bond parameters===============
> For bondtype: M1-Y1. There are 1 bond(s) in this type:
> The bond length(s) is(are): [2.0637]
> The average bond length is: 2.0637 Angstrom.
> The force constant(s) is(are): [np.float64(103.8)]
> The average force constant is: 103.8 Kcal*mol^-1*A^-2.
> For bondtype: M1-Y2. There are 1 bond(s) in this type:
> The bond length(s) is(are): [1.9964]
> The average bond length is: 1.9964 Angstrom.
> The force constant(s) is(are): [np.float64(144.8)]
> The average force constant is: 144.8 Kcal*mol^-1*A^-2.
> For bondtype: M1-Y3. There are 1 bond(s) in this type:
> The bond length(s) is(are): [2.311]
> The average bond length is: 2.311 Angstrom.
> The force constant(s) is(are): [np.float64(140.6)]
> The average force constant is: 140.6 Kcal*mol^-1*A^-2.
> For bondtype: M1-Y4. There are 1 bond(s) in this type:
> The bond length(s) is(are): [2.1182]
> The average bond length is: 2.1182 Angstrom.
> The force constant(s) is(are): [np.float64(87.7)]
> The average force constant is: 87.7 Kcal*mol^-1*A^-2.
> =======================Generate the angle parameters==============
> For angletype: M1-Y1-c1. There are 1 angle(s) in this type:
> The angle value(s) is(are): [134.7]
> The average angle value is: 134.7 Degree.
> The force constant(s) is(are): [np.float64(143.92)]
> The average force constant is: 143.92 Kcal*mol^-1*Rad^-2.
> Traceback (most recent call last):
> File
> "/mnt/storage_2/project_data/pl0251-02/bin/miniconda3/envs/AmberTools24/bin/MCPB.py",
>
> line 682, in <module>
> gene_by_QM_fitting_sem(smpdbf, ionids, xstru, stfpf, prefcdf,
> finfcdf,
>
> ~~~~~~~~~~~~~~~~~~~~~~^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
> fcfchkf, fclogf, g0x, scalef, bondfc_avg, anglefc_avg)
> ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
> File
> "/mnt/storage_2/project_data/pl0251-02/bin/miniconda3/envs/AmberTools24/lib/python3.13/site-packages/pymsmt/mcpb/gene_final_frcmod_file.py",
>
> line 835, in gene_by_QM_fitting_sem
> ang_para = avg_angle_para(misangat123, angvals, afconst)
> File
> "/mnt/storage_2/project_data/pl0251-02/bin/miniconda3/envs/AmberTools24/lib/python3.13/site-packages/pymsmt/mcpb/gene_final_frcmod_file.py",
>
> line 131, in avg_angle_para
> afconst = [round(i, 2) for i in afconst]
> ~~~~~^^^^^^
> TypeError: type numpy.complex128 doesn't define __round__ method
> """
>
> It seems that the program is attempting to round a value of type
> numpy.complex128, which leads to a TypeError since complex numbers
> cannot be directly rounded.
>
>
> For convenience, I’ve made all the relevant input files available in
> this GitHub repository:
> https://github.com/farmaceut/ERROR_MCPB.py.git
>
> Has anyone encountered a similar issue? Any suggestions or insights on
> how to resolve this would be greatly appreciated.
> Thank you in advance !
> -
> Maciej Spiegel, MPharm PhD
> assistant professor
>
> Department of Organic Chemistry and Pharmaceutical Technology,
> Faculty of Pharmacy, Wroclaw Medical University
> Borowska 211A, 50-556 Wroclaw, Poland
>
> _______________________________________________
> AMBER mailing list
> AMBER.ambermd.org
> http://lists.ambermd.org/mailman/listinfo/amber
>
_______________________________________________
AMBER mailing list
AMBER.ambermd.org
http://lists.ambermd.org/mailman/listinfo/amber
Received on Fri Aug 01 2025 - 09:30:03 PDT
