Download Cartesian Coordinates of

Hi forum, hi Markus ;o)

I am pretty bad with using NOMAD and APIs and stuff ;o)

My question is how do I get the xyz data that is used to, for example, render the images on the NOMAD website.

For example for this entry: tcMB0a7ifyVXBSgmnKj1-4w3XX21 (

How do I have to query in order to get out Cartesian coordinates of the format:

[atom type] [x coordinate] [y coordinate] [z coordinate]

… My question goes this direction:
This is a hybrid material of formula ABX3. We know what is A (molecule) and B (metal cation) and X (halogen anion), but we do not have a proper name for the molecule as physicists tend to forget that this is helpful.

From the Cartesian coordinates we could delete all lines that belong to B and X (in the case of this example: one line for “Sn” and three lines for “I”). What results is a coordinate file for the molecule “A”, in this case NH4.

These coordinates can be put into a tool to create a name/inchi/smiles code, e.g. something like this:

Many thanks!

Hi @caba!

I think currently the most straightforward way to get the coordinates (with minimal API knowledge :slight_smile: ) is to use the following type of a link:<entry_id>?path=<system path>&format=pdb

You could try pasting the following link into your browser window: Ideally this should start the file download immediately. It will return the first system defined in that entry, but you can target any other by changing the path.

You can find the details for this API route here, but briefly put this will return a pdb file for a specific system (indicated by path) in a specific entry (indicated by entry_id). Only PDB file format is currently supported, but more formats will be available in the future. These PDB files will contain the XYZ coordinates in plain text.

Let me know if this does not work out for you so we can think of another solution.

hi @laurih,
this excellent and solves my problem for now!
many thanks!

to some extend, we are enriching the information about this system, as we find out the name of the molecule in the material. is there a way to feed back information into NOMAD?

cheers, carsten

Good to hear!

Unfortunately, we do not support yet this kind of feedback loop where users could manually extend existing information in the archives. There are certain problems that come with this kind of user input (e.g. how to ascertain the validity of user-provided infomration, how to resolve contradictions between what is automatically detected vs. what users say etc.) that make this topic a bit tricky.

But we do have plans for automatically providing e.g. SMILES strings and InChI for molecules. The entry you are referring to is a bit more difficult because one needs to first separate the guest molecule from the host structure but we do have some ideas there as well.

We are doing this in a project where we try to link data from two data sources in order to answer queries using an ontology/knowledge graph. maybe we can exchange a bit on that at some point.