mlatom 3.18.0__tar.gz → 3.18.2__tar.gz

{mlatom-3.18.0/src/mlatom.egg-info → mlatom-3.18.2}/PKG-INFO

@@ -1,6 +1,6 @@
 Metadata-Version: 2.1
 Name: mlatom
-Version: 3.18.0
+Version: 3.18.2
 Summary: A Package for AI-enhanced computational chemistry
 Home-page: http://mlatom.com
 Author: Pavlo O. Dral
@@ -27,7 +27,7 @@ Requires-Dist: tqdm
 # About Program
 MLatom: a Package for Atomistic Simulations with Machine Learning
 
-Version 3.18.0
+Version 3.18.2
 
 **Official website**: http://mlatom.com/    
 **Documentation**: http://mlatom.com/docs/                     
@@ -56,7 +56,7 @@ When this Software or its derivatives are used in scientific publications, it sh
 * Pavlo O. Dral, Fuchun Ge, Yi-Fan Hou, Peikun Zheng, Yuxinxin Chen, Mario Barbatti, Olexandr Isayev, Cheng Wang, Bao-Xin Xue, Max Pinheiro Jr, Yuming Su, Yiheng Dai, Yangtao Chen, Lina Zhang, Shuang Zhang, Arif Ullah, Quanhao Zhang, Yanchi Ou. [MLatom 3: A Platform for Machine Learning-enhanced Computational Chemistry Simulations and Workflows](https://doi.org/10.1021/acs.jctc.3c01203).   *J. Chem. Theory Comput.* **2024**, *20*, 1193--1213.                                                                                               
 * Pavlo O. Dral, Fuchun Ge, Bao-Xin Xue, Yi-Fan Hou, Max Pinheiro Jr, Jianxing Huang, Mario Barbatti. [MLatom 2: An Integrative Platform for Atomistic Machine Learning](http://doi.org/10.1007/s41061-021-00339-5). *Top. Curr. Chem.* **2021**, *379*, 27.       
 * Pavlo O. Dral. [MLatom: A Program Package for Quantum Chemical Research Assisted by Machine Learning](http://dx.doi.org/10.1002/jcc.26004). *J. Comput. Chem.* **2019**, *40*, 2339--2347.                                                                                      
-* Pavlo O. Dral, Fuchun Ge, Yi-Fan Hou, Yuxinxin Chen, Peikun Zheng, Bao-Xin Xue, Mikolaj Martyka, Lina Zhang, Jakub Martinka, Quanhao Zhang, Xin-Yu Tong, Arif Ullah, Sebastian V. Pios, Vignesh B. Kumar, Yanchi Ou, Max Pinheiro Jr, Yuming Su, Yiheng Dai, Yangtao Chen, Shuang Zhang, Matheus O. Bispo, MLatom: A Package for Atomistic Simulations with Machine Learning,  version 3.18.0, Xiamen University, Xiamen, China, 2013-2025. [MLatom.com](http://mlatom.com).
+* Pavlo O. Dral, Fuchun Ge, Yi-Fan Hou, Yuxinxin Chen, Peikun Zheng, Bao-Xin Xue, Mikolaj Martyka, Lina Zhang, Jakub Martinka, Quanhao Zhang, Xin-Yu Tong, Arif Ullah, Sebastian V. Pios, Vignesh B. Kumar, Yanchi Ou, Max Pinheiro Jr, Yuming Su, Yiheng Dai, Yangtao Chen, Shuang Zhang, Matheus O. Bispo, MLatom: A Package for Atomistic Simulations with Machine Learning,  version 3.18.2, Xiamen University, Xiamen, China, 2013-2025. [MLatom.com](http://mlatom.com).
                                                                       
 The citations for MLatom's interfaces and features shall be eventually
 included too. See program output, header.py, ref.json, and [MLatom.com](http://mlatom.com).          

{mlatom-3.18.0 → mlatom-3.18.2}/README.md

@@ -1,7 +1,7 @@
 # About Program
 MLatom: a Package for Atomistic Simulations with Machine Learning
 
-Version 3.18.0
+Version 3.18.2
 
 **Official website**: http://mlatom.com/    
 **Documentation**: http://mlatom.com/docs/                     
@@ -30,7 +30,7 @@ When this Software or its derivatives are used in scientific publications, it sh
 * Pavlo O. Dral, Fuchun Ge, Yi-Fan Hou, Peikun Zheng, Yuxinxin Chen, Mario Barbatti, Olexandr Isayev, Cheng Wang, Bao-Xin Xue, Max Pinheiro Jr, Yuming Su, Yiheng Dai, Yangtao Chen, Lina Zhang, Shuang Zhang, Arif Ullah, Quanhao Zhang, Yanchi Ou. [MLatom 3: A Platform for Machine Learning-enhanced Computational Chemistry Simulations and Workflows](https://doi.org/10.1021/acs.jctc.3c01203).   *J. Chem. Theory Comput.* **2024**, *20*, 1193--1213.                                                                                               
 * Pavlo O. Dral, Fuchun Ge, Bao-Xin Xue, Yi-Fan Hou, Max Pinheiro Jr, Jianxing Huang, Mario Barbatti. [MLatom 2: An Integrative Platform for Atomistic Machine Learning](http://doi.org/10.1007/s41061-021-00339-5). *Top. Curr. Chem.* **2021**, *379*, 27.       
 * Pavlo O. Dral. [MLatom: A Program Package for Quantum Chemical Research Assisted by Machine Learning](http://dx.doi.org/10.1002/jcc.26004). *J. Comput. Chem.* **2019**, *40*, 2339--2347.                                                                                      
-* Pavlo O. Dral, Fuchun Ge, Yi-Fan Hou, Yuxinxin Chen, Peikun Zheng, Bao-Xin Xue, Mikolaj Martyka, Lina Zhang, Jakub Martinka, Quanhao Zhang, Xin-Yu Tong, Arif Ullah, Sebastian V. Pios, Vignesh B. Kumar, Yanchi Ou, Max Pinheiro Jr, Yuming Su, Yiheng Dai, Yangtao Chen, Shuang Zhang, Matheus O. Bispo, MLatom: A Package for Atomistic Simulations with Machine Learning,  version 3.18.0, Xiamen University, Xiamen, China, 2013-2025. [MLatom.com](http://mlatom.com).
+* Pavlo O. Dral, Fuchun Ge, Yi-Fan Hou, Yuxinxin Chen, Peikun Zheng, Bao-Xin Xue, Mikolaj Martyka, Lina Zhang, Jakub Martinka, Quanhao Zhang, Xin-Yu Tong, Arif Ullah, Sebastian V. Pios, Vignesh B. Kumar, Yanchi Ou, Max Pinheiro Jr, Yuming Su, Yiheng Dai, Yangtao Chen, Shuang Zhang, Matheus O. Bispo, MLatom: A Package for Atomistic Simulations with Machine Learning,  version 3.18.2, Xiamen University, Xiamen, China, 2013-2025. [MLatom.com](http://mlatom.com).
                                                                       
 The citations for MLatom's interfaces and features shall be eventually
 included too. See program output, header.py, ref.json, and [MLatom.com](http://mlatom.com).          

{mlatom-3.18.0 → mlatom-3.18.2}/setup.py

@@ -43,7 +43,7 @@ with open(os.path.join(here, 'README.md')) as f:
 
 setup(
     name = pkg_name,
-    version = "3.18.0",
+    version = "3.18.2",
     author = "Pavlo O. Dral",
     author_email = "admin@mlatom.com",
     license = 'MIT (modified)',

{mlatom-3.18.0 → mlatom-3.18.2}/src/mlatom/EEEDDDD.py

@@ -274,7 +274,7 @@ class EEEDDDD(model_cls.ml_model, model_cls.torch_model,model_cls.downloadable_m
             print(f"model saved in {model_file}")
 
     def _load_model(self, model_file: str):
-        model_dict = torch.load(model_file, map_location=self.device)
+        model_dict = torch.load(model_file, map_location=self.device, weights_only=False)
         self.species = model_dict['args'].pop('species')
         model = self._new_model(model_dict['args'])
         model.load_state_dict(model_dict['state'])

{mlatom-3.18.0 → mlatom-3.18.2}/src/mlatom/MLatom.py

@@ -4,7 +4,7 @@
   !---------------------------------------------------------------------------!
   !                                                                           !
   !     MLatom: a Package for Atomistic Simulations with Machine Learning     !
-  !                             MLatom 3.18.0                                 !
+  !                             MLatom 3.18.2                                 !
   !                                   @                                       !
   !                 Xiamen Atomistic Computing Suite (XACS)                   !
   !                                                                           !
@@ -38,7 +38,7 @@
   ! Max Pinheiro Jr, Yuming Su, Yiheng Dai, Yangtao Chen, Shuang Zhang,       !
   ! Matheus O. Bispo                                                          !
   ! MLatom: A Package for Atomistic Simulations with Machine Learning,        !
-  ! version 3.18.0, Xiamen University, Xiamen, China, 2013-2025.              !
+  ! version 3.18.2, Xiamen University, Xiamen, China, 2013-2025.              !
   !                                                                           !
   ! The citations for MLatom's interfaces and features shall be eventually    !
   ! included too. See header.py, ref.json and http://mlatom.com.              !

{mlatom-3.18.0 → mlatom-3.18.2}/src/mlatom/__init__.py

@@ -5,11 +5,11 @@ from .simulations import optimize_geometry, irc, freq, thermochemistry, md, gene
 from .data import atom, molecule, molecular_database
 from .models import kreg, ani, msani
 from .models import methods, aiqm1, aiqm2, dens, ani_methods, aimnet2_methods, gaussian_methods, pyscf_methods, orca_methods, turbomole_methods, mndo_methods, sparrow_methods, xtb_methods, dftbplus_methods, columbus_methods, dftd3_methods, dftd4_methods
+from .models import MDtrajNet
 from . import al_utils
 from .al import al
 from .gap_md import gap_md, gap_model
 # add-ons
 from .models import uaiqm, omnip2x, vecmsani
-from .EEEDDDD import EEEDDDD as MDtrajNet
 
-__version__ = '3.18.0'
+__version__ = '3.18.2'

{mlatom-3.18.0 → mlatom-3.18.2}/src/mlatom/al.py

@@ -8,6 +8,11 @@ import random
 import timeit
 import copy
 
+# check numpy version
+np_ver = np.__version__.split()[0]
+if np_ver == '1': np_cfloat = np.cfloat
+else: np_cfloat = np.complex128
+
 class al():
     '''
     Active learning procedure 
@@ -335,7 +340,8 @@ class al():
             elif type(dd[key]) == np.int16: dd[key] = dd[key].item()
             elif type(dd[key]) == np.int32: dd[key] = dd[key].item()
             elif type(dd[key]) == np.int64: dd[key] = dd[key].item()
-            elif type(dd[key]) == np.cfloat: dd[key] = dd[key].item()
+            # elif type(dd[key]) == np.cfloat: dd[key] = dd[key].item()
+            elif type(dd[key]) == np_cfloat: dd[key] = dd[key].item()
             elif type(dd[key]) == list:
                 for ii in range(len(dd[key])):
                     if type(dd[key][ii]) == list:

{mlatom-3.18.0 → mlatom-3.18.2}/src/mlatom/models.py

@@ -588,6 +588,13 @@ def mace(**kwargs):
     from .interfaces.mace_interface import mace
     return mace(**kwargs)
 
+def MDtrajNet(**kwargs):
+    '''
+    Returns an EEEDDDD model object (see :class:`mlatom.EEEDDDD.EEEDDDD`).
+    '''
+    from .EEEDDDD import EEEDDDD 
+    return EEEDDDD(**kwargs)
+
 from .addons.uaiqm.uaiqm             import uaiqm
 from .addons.omnip2x.omnip2x         import omnip2x
 from .addons.omnip2x.vecmsani        import vecmsani

{mlatom-3.18.0 → mlatom-3.18.2}/src/mlatom/namd.py

@@ -345,7 +345,7 @@ class surface_hopping_md():
                 self.current_state = hopping_probabilities.index(max_prob)
 
                 # fssh/lzsh/znsh: rescale_velocity; change en grad in molecular_trajectory; change ekin etot
-                mol_istep_plus1 = self.molecular_trajectory.steps[istep+1].molecule
+                mol_istep_plus1 = self.molecular_trajectory.steps[-2].molecule
                 kinetic_energy_change = mol_istep_plus1.electronic_states[self.current_state].energy - mol_istep_plus1.electronic_states[self.initial_state].energy
                 vector = None
                 if self.rescale_velocity_direction.casefold() in ['nacv', 'nacs']:
@@ -359,7 +359,7 @@ class surface_hopping_md():
                     kinetic_energy_change=kinetic_energy_change,
                     vector=vector,
                     if_not_enough_kinetic_energy=self.insufficient_energy_action)
-                self.change_properties_of_hopping_step(step=istep+1)
+                self.change_properties_of_hopping_step()
                 if self.hopping_algorithm == 'LZBL' or self.hopping_algorithm == 'FSSH':
                     del self.molecular_trajectory.steps[-1]
                     one_step_propagation = True
@@ -401,12 +401,22 @@ class surface_hopping_md():
                         temp_traj_dump.steps.append(self.molecular_trajectory.steps[-1])
                         if self.reduce_memory_usage:
                             temp_traj = data.molecular_trajectory()
-                            temp_traj.steps.append(self.molecular_trajectory.steps[-2])
-                            temp_traj.steps.append(self.molecular_trajectory.steps[-1])
-                            del self.molecular_trajectory.steps[-2:] 
-                            self.molecular_trajectory.dump(filename=self.filename, format=self.format)
-                            del self.molecular_trajectory
-                            self.molecular_trajectory = temp_traj
+                            if self.hopping_algorithm == 'LZBL':
+                                temp_traj.steps.append(self.molecular_trajectory.steps[-2])
+                                temp_traj.steps.append(self.molecular_trajectory.steps[-1])
+                                del self.molecular_trajectory.steps[-2:] 
+                                self.molecular_trajectory.dump(filename=self.filename, format=self.format)
+                                del self.molecular_trajectory
+                                self.molecular_trajectory = temp_traj
+                            if self.hopping_algorithm == 'FSSH':
+                                temp_traj.steps.append(self.molecular_trajectory.steps[-4])
+                                temp_traj.steps.append(self.molecular_trajectory.steps[-3])
+                                temp_traj.steps.append(self.molecular_trajectory.steps[-2])
+                                temp_traj.steps.append(self.molecular_trajectory.steps[-1])
+                                del self.molecular_trajectory.steps[-4:] 
+                                self.molecular_trajectory.dump(filename=self.filename, format=self.format)
+                                del self.molecular_trajectory
+                                self.molecular_trajectory = temp_traj
                         else:
                             temp_traj_dump.dump(filename=self.filename, format=self.format)
                     elif self.format == 'json':
@@ -504,27 +514,27 @@ class surface_hopping_md():
         n_substeps = int(self.time_step / self.time_step_tdse)
 
         # Load energies and interpolate
-        energies = [np.array([self.molecular_trajectory.steps[istep+1].molecule.electronic_states[iState].energy for iState in range(self.nstates)]), np.array([self.molecular_trajectory.steps[istep].molecule.electronic_states[stat].energy for stat in range(self.nstates)])]
+        energies = [np.array([self.molecular_trajectory.steps[-2].molecule.electronic_states[iState].energy for iState in range(self.nstates)]), np.array([self.molecular_trajectory.steps[-3].molecule.electronic_states[stat].energy for stat in range(self.nstates)])]
         if istep == 1:
-            energies.append(np.array([self.molecular_trajectory.steps[istep-1].molecule.electronic_states[iState].energy for iState in range(self.nstates)]))
+            energies.append(np.array([self.molecular_trajectory.steps[-4].molecule.electronic_states[iState].energy for iState in range(self.nstates)]))
         elif istep > 1:
-            energies.append(np.array([self.molecular_trajectory.steps[istep-1].molecule.electronic_states[iState].energy for iState in range(self.nstates)]))
-            energies.append(np.array([self.molecular_trajectory.steps[istep-2].molecule.electronic_states[iState].energy for iState in range(self.nstates)]))
+            energies.append(np.array([self.molecular_trajectory.steps[-4].molecule.electronic_states[iState].energy for iState in range(self.nstates)]))
+            energies.append(np.array([self.molecular_trajectory.steps[-5].molecule.electronic_states[iState].energy for iState in range(self.nstates)]))
         energies_interpolated = [list(ienergy) for ienergy in zip(*self.interpolate_energy(energies, n_substeps))]
 
         # Load velocities
-        velocity_initial = np.array(self.molecular_trajectory.steps[istep].molecule.get_xyz_vectorial_properties('xyz_velocities'))*constants.Angstrom2Bohr/constants.fs2au
-        velocity_final = np.array(self.molecular_trajectory.steps[istep+1].molecule.get_xyz_vectorial_properties('xyz_velocities'))*constants.Angstrom2Bohr/constants.fs2au
+        velocity_initial = np.array(self.molecular_trajectory.steps[-3].molecule.get_xyz_vectorial_properties('xyz_velocities'))*constants.Angstrom2Bohr/constants.fs2au
+        velocity_final = np.array(self.molecular_trajectory.steps[-2].molecule.get_xyz_vectorial_properties('xyz_velocities'))*constants.Angstrom2Bohr/constants.fs2au
 
         # Load NACs
-        nac_initial = np.array(self.molecular_trajectory.steps[istep].molecule.nacv)/constants.Angstrom2Bohr
-        nac_final = np.array(self.molecular_trajectory.steps[istep+1].molecule.nacv)/constants.Angstrom2Bohr
+        nac_initial = np.array(self.molecular_trajectory.steps[-3].molecule.nacv)/constants.Angstrom2Bohr
+        nac_final = np.array(self.molecular_trajectory.steps[-2].molecule.nacv)/constants.Angstrom2Bohr
         for iState in range(self.nstates):
             for jState in range(iState):
                 # Phase alignment within trajectory based on the dot product of NAC(t) and NAC(t+dt)
                 if np.vdot(nac_initial[iState][jState], nac_final[iState][jState]) < 0:
-                    self.molecular_trajectory.steps[istep+1].molecule.nacv[iState][jState] *= -1
-                    self.molecular_trajectory.steps[istep+1].molecule.nacv[jState][iState] *= -1
+                    self.molecular_trajectory.steps[-2].molecule.nacv[iState][jState] *= -1
+                    self.molecular_trajectory.steps[-2].molecule.nacv[jState][iState] *= -1
                     nac_final[iState][jState] *= -1
                     nac_final[jState][iState] *= -1
 
@@ -537,13 +547,13 @@ class surface_hopping_md():
             ss_c = [self.molecular_trajectory.steps[istep].state_coefficients]
             ss_cdot = [self.coeff_dot(ss_c[-1], ss_dt[-1], ss_ph[-1])]
         else:
-            ss_E = self.molecular_trajectory.steps[istep-1].substep_potential_energy[-4:]
-            ss_v = self.molecular_trajectory.steps[istep-1].substep_velocities[-4:]
-            ss_NAC = self.molecular_trajectory.steps[istep-1].substep_nonadiabatic_coupling_vectors[-4:]
-            ss_dt = self.molecular_trajectory.steps[istep-1].substep_time_derivative_coupling[-4:]
-            ss_c = self.molecular_trajectory.steps[istep-1].substep_state_coefficients[-4:]
-            ss_cdot = self.molecular_trajectory.steps[istep-1].substep_state_coefficients_dot[-4:]
-            ss_ph = self.molecular_trajectory.steps[istep-1].substep_phase[-4:]
+            ss_E = self.molecular_trajectory.steps[-4].substep_potential_energy[-4:]
+            ss_v = self.molecular_trajectory.steps[-4].substep_velocities[-4:]
+            ss_NAC = self.molecular_trajectory.steps[-4].substep_nonadiabatic_coupling_vectors[-4:]
+            ss_dt = self.molecular_trajectory.steps[-4].substep_time_derivative_coupling[-4:]
+            ss_c = self.molecular_trajectory.steps[-4].substep_state_coefficients[-4:]
+            ss_cdot = self.molecular_trajectory.steps[-4].substep_state_coefficients_dot[-4:]
+            ss_ph = self.molecular_trajectory.steps[-4].substep_phase[-4:]
         ss_random_numbers = []
         ss_hopping_probabilities = []
 
@@ -613,12 +623,12 @@ class surface_hopping_md():
                 if max(hopping_probabilities) > random_number:
                     hop_occured = True
 
-            ss_c[-1] = self.decoherence_correction(istep, ss_c[-1])
+            ss_c[-1] = self.decoherence_correction(-2, ss_c[-1])
             
             ss_hopping_probabilities.append(hopping_probabilities)
             ss_random_numbers.append(random_number)
 
-        self.molecular_trajectory.steps[istep+1].state_coefficients = ss_c[-1]
+        self.molecular_trajectory.steps[-2].state_coefficients = ss_c[-1]
 
         # Saving substep info: last element of istep is the same as first element of istep+1
         if istep == 0:
@@ -630,15 +640,15 @@ class surface_hopping_md():
             self.molecular_trajectory.steps[istep].substep_state_coefficients_dot = ss_cdot
             self.molecular_trajectory.steps[istep].substep_phase = ss_ph
         else:
-            self.molecular_trajectory.steps[istep].substep_potential_energy = ss_E[3:]
-            self.molecular_trajectory.steps[istep].substep_velocities = ss_v[3:]
-            self.molecular_trajectory.steps[istep].substep_nonadiabatic_coupling_vectors = ss_NAC[3:]
-            self.molecular_trajectory.steps[istep].substep_time_derivative_coupling = ss_dt[3:]
-            self.molecular_trajectory.steps[istep].substep_state_coefficients = ss_c[3:]
-            self.molecular_trajectory.steps[istep].substep_state_coefficients_dot = ss_cdot[3:]
-            self.molecular_trajectory.steps[istep].substep_phase = ss_ph[3:]
-        self.molecular_trajectory.steps[istep].substep_random_numbers = ss_random_numbers
-        self.molecular_trajectory.steps[istep].substep_hopping_probabilities = ss_hopping_probabilities
+            self.molecular_trajectory.steps[-3].substep_potential_energy = ss_E[3:]
+            self.molecular_trajectory.steps[-3].substep_velocities = ss_v[3:]
+            self.molecular_trajectory.steps[-3].substep_nonadiabatic_coupling_vectors = ss_NAC[3:]
+            self.molecular_trajectory.steps[-3].substep_time_derivative_coupling = ss_dt[3:]
+            self.molecular_trajectory.steps[-3].substep_state_coefficients = ss_c[3:]
+            self.molecular_trajectory.steps[-3].substep_state_coefficients_dot = ss_cdot[3:]
+            self.molecular_trajectory.steps[-3].substep_phase = ss_ph[3:]
+        self.molecular_trajectory.steps[-3].substep_random_numbers = ss_random_numbers
+        self.molecular_trajectory.steps[-3].substep_hopping_probabilities = ss_hopping_probabilities
 
         return hopping_probabilities, random_number
 
@@ -742,7 +752,7 @@ class surface_hopping_md():
             coeff[self.current_state] *= np.sqrt((1-coeff_sum)/(np.abs(coeff[self.current_state])**2))
         return coeff
     
-    def change_properties_of_hopping_step(self, step):
+    def change_properties_of_hopping_step(self):
         new_epot = self.molecular_trajectory.steps[-2].molecule.electronic_states[self.current_state].energy
         self.molecular_trajectory.steps[-2].molecule.energy = new_epot
         # for atom in self.molecular_trajectory.steps[step].molecule.atoms:

{mlatom-3.18.0 → mlatom-3.18.2/src/mlatom.egg-info}/PKG-INFO

@@ -1,6 +1,6 @@
 Metadata-Version: 2.1
 Name: mlatom
-Version: 3.18.0
+Version: 3.18.2
 Summary: A Package for AI-enhanced computational chemistry
 Home-page: http://mlatom.com
 Author: Pavlo O. Dral
@@ -27,7 +27,7 @@ Requires-Dist: tqdm
 # About Program
 MLatom: a Package for Atomistic Simulations with Machine Learning
 
-Version 3.18.0
+Version 3.18.2
 
 **Official website**: http://mlatom.com/    
 **Documentation**: http://mlatom.com/docs/                     
@@ -56,7 +56,7 @@ When this Software or its derivatives are used in scientific publications, it sh
 * Pavlo O. Dral, Fuchun Ge, Yi-Fan Hou, Peikun Zheng, Yuxinxin Chen, Mario Barbatti, Olexandr Isayev, Cheng Wang, Bao-Xin Xue, Max Pinheiro Jr, Yuming Su, Yiheng Dai, Yangtao Chen, Lina Zhang, Shuang Zhang, Arif Ullah, Quanhao Zhang, Yanchi Ou. [MLatom 3: A Platform for Machine Learning-enhanced Computational Chemistry Simulations and Workflows](https://doi.org/10.1021/acs.jctc.3c01203).   *J. Chem. Theory Comput.* **2024**, *20*, 1193--1213.                                                                                               
 * Pavlo O. Dral, Fuchun Ge, Bao-Xin Xue, Yi-Fan Hou, Max Pinheiro Jr, Jianxing Huang, Mario Barbatti. [MLatom 2: An Integrative Platform for Atomistic Machine Learning](http://doi.org/10.1007/s41061-021-00339-5). *Top. Curr. Chem.* **2021**, *379*, 27.       
 * Pavlo O. Dral. [MLatom: A Program Package for Quantum Chemical Research Assisted by Machine Learning](http://dx.doi.org/10.1002/jcc.26004). *J. Comput. Chem.* **2019**, *40*, 2339--2347.                                                                                      
-* Pavlo O. Dral, Fuchun Ge, Yi-Fan Hou, Yuxinxin Chen, Peikun Zheng, Bao-Xin Xue, Mikolaj Martyka, Lina Zhang, Jakub Martinka, Quanhao Zhang, Xin-Yu Tong, Arif Ullah, Sebastian V. Pios, Vignesh B. Kumar, Yanchi Ou, Max Pinheiro Jr, Yuming Su, Yiheng Dai, Yangtao Chen, Shuang Zhang, Matheus O. Bispo, MLatom: A Package for Atomistic Simulations with Machine Learning,  version 3.18.0, Xiamen University, Xiamen, China, 2013-2025. [MLatom.com](http://mlatom.com).
+* Pavlo O. Dral, Fuchun Ge, Yi-Fan Hou, Yuxinxin Chen, Peikun Zheng, Bao-Xin Xue, Mikolaj Martyka, Lina Zhang, Jakub Martinka, Quanhao Zhang, Xin-Yu Tong, Arif Ullah, Sebastian V. Pios, Vignesh B. Kumar, Yanchi Ou, Max Pinheiro Jr, Yuming Su, Yiheng Dai, Yangtao Chen, Shuang Zhang, Matheus O. Bispo, MLatom: A Package for Atomistic Simulations with Machine Learning,  version 3.18.2, Xiamen University, Xiamen, China, 2013-2025. [MLatom.com](http://mlatom.com).
                                                                       
 The citations for MLatom's interfaces and features shall be eventually
 included too. See program output, header.py, ref.json, and [MLatom.com](http://mlatom.com).