I get the error below when trying to import a pdb-file, online or localy.

I have followed the steps in the instalation guide: I run Blender as an administrator on Windows 10, and after installing the add-on, i click the button to install Atomium. I get a confirmation that Atomium is successfully installed.

In the MolecularNodes panel i click the "Download" / "Open" button and get the error below. I assume the it is because the Atomium library is not properly installed.

Python: Traceback (most recent call last):
  File "C:\Users\Jonas\AppData\Roaming\Blender Foundation\Blender\3.1\scripts\addons\molecular_nodes\__init__.py", line 1401, in execute
    sna_func_import_structure_EE5B6(bpy.context.scene.sna_pdb_code,
  File "C:\Users\Jonas\AppData\Roaming\Blender Foundation\Blender\3.1\scripts\addons\molecular_nodes\__init__.py", line 1134, in sna_func_import_structure_EE5B6
    pdb = atomium.fetch(pdb_id)
UnboundLocalError: local variable 'atomium' referenced before assignment

location: <unknown location>:-1

I am very curious to try this amazing add-on, hope that you can help resolve the issue.

hello, when i render my model, some problem appear.

1. i use chimera to generate small molecue dymamics file (.pdb)
2. render lost atom -br see 2d structure reference：https://pubchem.ncbi.nlm.nih.gov/compound/66789061
3. model file and render results plz see [https://github.com/wanjiangbo1985/mdynamics]

thank u for your great work

wanjiangbo

Hi Brady,

Thanks for developing such an awesome node for the scientific community. I'm just getting taste of the depth of Blender and your node while watching your tutorials to reproduce. I found most of functionalities and logical flows are very intuitive and I believe I can learn ton more by just simply connecting nodes here and there.

Unfortunately, I'm having a bit of difficulty with coloring ribbon/surface/atom by chain ID manually rather than by random value node. For example, in ChimeraX, I would simply do "color /D #hexColor target c" for cartoon coloring with a specific color on a specific chain but I can't figure out how i can do the equivalent thing in this context.

Can you please give me some hints if possible?

Thank you so much.

best, heejong

I am trying to colour a dimeric protein based on chain id, but the chain ids in the pdb seem to get mangled during the import. Or maybe I am misunderstanding something. My setup is pretty simple (see image below). I take the chain_number and connect it to a Map Range node, mapping 1-2 to 0-1 (tried different ranges), which is then passed to a ColorRamp node. The colour from the ramp is passed to all elements of a MOL_style_colour node, which in turn connects to the geometry output. However, the resulting colouring seems more or less random, albeit anticorrelated between the two chains (see image). The chain labels for the two monomers in the pdb file are C and D. Pymol seems to handle this fine (third image), so I don't think the pdb file is broken.

Is this a bug? I will try and see if I can spot something in the source code and update if I find something suspicious.

I'm trying to load a PDB from a MD simulation from a folder. It has no 4-character code and all the times i tried to, it failed.

Is there a way to load a "custom" pdb or only from PDBDatabank?

Thank you!

@orbeckst Thanks for the installation guide! That's really helpful and I got atomium and MDA linking to work(now both buttons are grey). When I was trying to load the pdb and xtc, I still got the following error. It seems to be something from blender .

Originally posted by @ChenouZhang in https://github.com/BradyAJohnston/MolecularNodes/issues/24#issuecomment-1174447127

Hello Brady :)

Long time ago I made a video on VMD of my Norovirus Capsid, and I thought that it could be a good test for your tool (Since I have 600K+ atom in the PDB). Here it is (just 1 state) morph_state001-capsid.zip

The problem is that only one part seems to be read since the numbering stop at 99999 (cf. Screenshot 1). I tried different format (convertion with Pymol to mmtf, mmcif, cif, pdb.gz, cif.gz) and it fails for other format (But here the problem seems to come from the atomium library for reading mmtf or mmcif when the output comme from Pymol maybe ?)

For this issue maybe the problem come from the atomium library itself ? But I wanted to repport it since with this new render system it seems possible to load very big systems, and unfortunatly the "assembly" coming from the PDB is splitted into different models (so your addon think it's a trajectory)

Anyway, here's some extra information ⤵️ and thank you again for this already great piece of work =D Thibault.

Screenshot 1:

traceback for loading when it's mmtf

Python: Traceback (most recent call last):
  File "C:\Users\tubia\AppData\Roaming\Blender Foundation\Blender\3.1\scripts\addons\molecular_nodes\__init__.py", line 80, in execute
    sna_func_import_structure_EE5B6(bpy.context.scene.sna_pdb_code,
  File "C:\Users\tubia\AppData\Roaming\Blender Foundation\Blender\3.1\scripts\addons\molecular_nodes\__init__.py", line 1137, in sna_func_import_structure_EE5B6
    pdb = atomium.open(pdb_path)
  File "C:\Users\tubia\AppData\Roaming\Python\Python310\site-packages\atomium\utilities.py", line 41, in open
    return parse_string(filestring, path, *args, **kwargs)
  File "C:\Users\tubia\AppData\Roaming\Python\Python310\site-packages\atomium\utilities.py", line 122, in parse_string
    parsed = file_func(filestring)
  File "C:\Users\tubia\AppData\Roaming\Python\Python310\site-packages\atomium\mmtf.py", line 18, in mmtf_bytes_to_mmtf_dict
    return decode_dict(raw)
  File "C:\Users\tubia\AppData\Roaming\Python\Python310\site-packages\atomium\mmtf.py", line 40, in decode_dict
    new_value = [decode_dict(x) for x in new_value]
  File "C:\Users\tubia\AppData\Roaming\Python\Python310\site-packages\atomium\mmtf.py", line 40, in <listcomp>
    new_value = [decode_dict(x) for x in new_value]
  File "C:\Users\tubia\AppData\Roaming\Python\Python310\site-packages\atomium\mmtf.py", line 37, in decode_dict
    new_value = parse_binary_field(new_value)
  File "C:\Users\tubia\AppData\Roaming\Python\Python310\site-packages\atomium\mmtf.py", line 56, in parse_binary_field
    codec, length, params = struct.unpack(">iii", b[:12])
struct.error: unpack requires a buffer of 12 bytes

location: <unknown location>:-1

Traceback for loading when it's mmcif

Python: Traceback (most recent call last):
  File "C:\Users\tubia\AppData\Roaming\Python\Python310\site-packages\atomium\pdb.py", line 506, in add_atom_to_non_polymer
    ] = atom_line_to_dict(line, aniso_dict)
  File "C:\Users\tubia\AppData\Roaming\Python\Python310\site-packages\atomium\pdb.py", line 530, in atom_line_to_dict
    a["x"] = float(line[30:38].strip())
ValueError: could not convert string to float: '9 ? -57.'

During handling of the above exception, another exception occurred:

Traceback (most recent call last):
  File "C:\Users\tubia\AppData\Roaming\Blender Foundation\Blender\3.1\scripts\addons\molecular_nodes\__init__.py", line 80, in execute
    sna_func_import_structure_EE5B6(bpy.context.scene.sna_pdb_code,
  File "C:\Users\tubia\AppData\Roaming\Blender Foundation\Blender\3.1\scripts\addons\molecular_nodes\__init__.py", line 1137, in sna_func_import_structure_EE5B6
    pdb = atomium.open(pdb_path)
  File "C:\Users\tubia\AppData\Roaming\Python\Python310\site-packages\atomium\utilities.py", line 41, in open
    return parse_string(filestring, path, *args, **kwargs)
  File "C:\Users\tubia\AppData\Roaming\Python\Python310\site-packages\atomium\utilities.py", line 124, in parse_string
    parsed = data_func(parsed)
  File "C:\Users\tubia\AppData\Roaming\Python\Python310\site-packages\atomium\pdb.py", line 86, in pdb_dict_to_data_dict
    update_models_list(pdb_dict, data_dict)
  File "C:\Users\tubia\AppData\Roaming\Python\Python310\site-packages\atomium\pdb.py", line 158, in update_models_list
    add_atom_to_non_polymer(line, model, res_id, aniso, full_names)
  File "C:\Users\tubia\AppData\Roaming\Python\Python310\site-packages\atomium\pdb.py", line 512, in add_atom_to_non_polymer
    "atoms": {int(line[6:11]): atom_line_to_dict(line, aniso_dict)}
  File "C:\Users\tubia\AppData\Roaming\Python\Python310\site-packages\atomium\pdb.py", line 530, in atom_line_to_dict
    a["x"] = float(line[30:38].strip())
ValueError: could not convert string to float: '9 ? -57.'

location: <unknown location>:-1

Currently there is no way to select, delete or distinguish water molecules in the structures that are imported.

Having a is_water property would fix this issue.

Hi!

I installed MolecularNodes, both the add-on and the atomium installation was said to be fine. However, when I tried to load a pdb file (even from pdb or file), I get the following error: Blender version: 2.93.4, add-on version: 0.4.0

What could the the issue?

To enable features like secondary structure detection and cartoon representations of proteins and nucleic acids, having the atom name field as an attribute inside of MN / GN.

For some atoms this is currently hardcoded like is_alpha_carbon attribute field for alpha carbons, but this should be used more broadly so selections of particular parts of particular amino acids and molecules would be possible.

The question would be how to numerically encode this information. For example, an array of atom names ['C', 'O', 'C1, 'C'] could be alphabetically ranked and numerically encoded as [0, 2, 1, 0] for GN to understand it.

From looking into it briefly, I haven't been able to find a definitive list of which atom names are pre-defined and possible to be used. I assume that there is a common list for the 20 natural amino acids and the different nucleic acids, but then it could potentially be defined by users for more edge cases?

Looking at an example structure 6n2y:

import MolecularNodes as mn
import numpy as np
mol, file = mn.load.open_structure_rcsb('6n2y')
np.unique(mol.atom_name)

array(['C', "C1'", 'C2', "C2'", "C3'", 'C4', "C4'", 'C5', "C5'", 'C6',
       'C8', 'CA', 'CB', 'CD', 'CD1', 'CD2', 'CE', 'CE1', 'CE2', 'CE3',
       'CG', 'CG1', 'CG2', 'CH2', 'CZ', 'CZ2', 'CZ3', 'MG', 'N', 'N1',
       'N3', 'N6', 'N7', 'N9', 'ND1', 'ND2', 'NE', 'NE1', 'NE2', 'NH1',
       'NH2', 'NZ', 'O', 'O1', 'O1A', 'O1B', 'O1G', 'O2', "O2'", 'O2A',
       'O2B', 'O2G', 'O3', "O3'", 'O3A', 'O3B', 'O3G', 'O4', "O4'", "O5'",
       'OD1', 'OD2', 'OE1', 'OE2', 'OG', 'OG1', 'OH', 'P', 'PA', 'PB',
       'PG', 'SD', 'SG'], dtype='<U6')

One could use a set of some amount to generate a list to work from. If you then have that list - how do you order it? So far I have just been encoding things in alphabetical order which I believe works well for amino acids. For these atom names however Having O so far down the list seems wrong, and I feel like the backbone atoms should be listed first. This could potentially make it easier to work with from a human perspective when having to manually define nodes inside of Blender to work with the numerical representations of these atom names. If you order the backbone atom names higher, should you also apply a level of order to the other atom names as well? If you end up with atom names which weren't accounted for, then where will they fall within the ordering?

I would love any feedback or comments on the matter. Unfortunately, this will have to be somewhat hardcoded inside the nodes themselves, so updating will be tricky and a logical implementation the first time will be ideal.

I am struggling to get the AA_sequence_number attribute to work with MOL_style_ribbon. This is problematic if I want to use the sequence ID for something other than colouring, for example, the emission strength of the material.

I have made an example below, which I tweak to show the behaviour with different node wiring. It is a dimer with chain_number 1 and 2, and each monomer has 129 aa, which are properly numbered in the pdb file. I try to store AA_sequence_number after mapping to the range 0–1 in a named property 'bzz'.

Using the chain_number instead of AA_sequence_number works fine. The chains get different colour by passing bzz through a color_ramp in the material node tree.

Mapping and storing AA_sequence_number doesn't work however.

I know that AA_sequence_number is correct before the MOL_style_ribbon. I can for example use it to set the colour of C atoms and use the Color attribute for colouring the material. Note however that this doesn't solve the general problem, because I might want to use the attribute for something else, in addition to some clever colouring based on other features.

I can also access chain_number directly from the materials node tree, but AA_sequence_number is a dead end.

I have inspected the geometry in the spreadsheet editor via a viewer node, and indeed the points are rich with attributes before the MOL_style_ribbon node, but the geometry that comes out have lost most of the attributes. I don't see a good reason why chain_number can be passed on to the ribbons but not AA_sequence_number.

While previously using atomium for parsing, b_factor on a per-model basis was available.

With the switch to Biotite, this is currently not available and b_factor is now taken from the first model and used for all other models as well.

The issue is raised here and a fix will hopefully be coming when new functionality is added Biotite.

Attempt for a General Support for mmCIF biological assembly. Some code extract from ePMV to create the instance mesh. Handle cellpack cif files. I dont know if there is a performance difference between geometrynode and Instancing on Faces using the regular interface. Two examples : 2PLV and a cellpack model.

Biotite can read other more chemically-oriented file types. Need to add specific support for it inside of MN. Will require some tweaking as they won't contain the same types of attributes that protein-oriented file types encode.

Hi, I tried to install Molecular Nodes 2.0.2 with Blender 3.4.0 on M1 Max Mac Ventura 13.0. After installing the .zip file, there is a button "install packages" in the Molecular Nodes panel. When I click it, I see a beachball and after a while it says "Failed to install required packages. Please check log file: ..." That log file just contains this:

Installer Started: 2022-12-17 15:55:53.117289

################################### Installer finished: 2022-12-17 15:56:24.477652 ###################################

Any idea what might be the problem?

Thanks, Gregor