Fetching data with BCIF instead of MMTF?

[dominiquesydow]

Hello :)

With MMTF replaced now with BCIF (see announcement here), do you know the equivalent to the following code for the BCIF API?

import mmtf
mmtf.fetch(pdbid)

Thanks!

[piehld]

Hi @dominiquesydow,

There are a few options that I can share:

• The way I actually recommend working with BCIF (both reading and writing) is using our rcsb.utils.io package, which is a wrapper for py-mmcif, since it lets you easily switch between format types without having to import different modules. The README.md there offers some general examples on its usage: https://github.com/rcsb/py-rcsb_utils_io (in which, for the case of reading BCIF, you would use fmt="bcif").
• If you specifically want to use py-mmcif, though, you can take a look at the testBinaryCifReader.py file, e.g.: https://github.com/rcsb/py-mmcif/blob/f7c5f1498d687115a72618d79f176b7af26f5ea7/mmcif/tests/testBinaryCifReader.py#L91-L92. (Also note that there is a testBinaryCifWriter.py file next to that one.)

Please let me know if you'd like more details or need any help with either of the above options. I can certainly elaborate as needed. Hope this helps as a start at least!

[dominiquesydow]

Hi @piehld,

Many thanks for your quick help!
I will try this out later this week and come back here if I get stuck.

Leaving it up to you if you'd like to close this issue as solved already.
Leaving it open could be nice for visibility in case others also still use the mmtf package and are in search of alternatives.

Many thanks again!

[piehld]

Hi @dominiquesydow,

It has been a while, but our team just presented a webinar on using these python packages for working with mmCIF and BCIF data, the presentation/slides of which will be made available in the coming weeks/months. However, for the meantime, I prepared a Jupyter notebook that I demo'ed for the webinar which I wanted to share with you now, in case you find it useful: https://github.com/rcsb/rcsb-training-resources/tree/master/training-events/2024/utilizing-binary-cif

When you open up that notebook, you'll see an "Open in Colab" badge at the top of it. I would recommend clicking that so you can run through the notebook on your own and test it out. I hope the comments in that file are clear. Please also stay tuned into our rcsb.utils.io and mmcif repos/packages to follow along with updates and enhancements that we will be introducing over the next few months.

Hope this offers additional help to you and anyone else who comes across this issue :).

Dennis

[llevoc]

Not sure that this is the correct forum but here goes:
Structure.from_pdbid(pdb_id) uses mmtf.fetch, which is firewall restricted. While reading cif files does not use fetch, the additional data processing features available in mmtf.fetch are lost. The post above by dominiquesydow of a colab session is useful. However,
duplicating the output of mmtf.fetch is lacking. Any chance anyone here could address this? Note that talktorials T008 and T015 use mmtf.fetch, and in my hands cannot be completed, since I have not yet found a workaround. Please note that I am NOT a python expert.
Thanks,

[pwrose]

mmtf-python does not support BCIF and the distribution of files in the MMTF format was discontinued. Please ask the authors of the tutorial for help with their colab notebooks.

…

On Mon, Sep 29, 2025 at 12:57 PM llevoc ***@***.***> wrote: *llevoc* left a comment (rcsb/mmtf-python#63) <#63 (comment)> Not sure that this is the correct forum but here goes: Structure.from_pdbid(pdb_id) uses mmtf.fetch, which is firewall restricted. While reading cif files does not use fetch, the additional data processing features available in mmtf.fetch are lost. The post above by dominiquesydow of a colab session is useful. However, duplicating the output of mmtf.fetch is lacking. Any chance anyone here could address this? Note that talktorials T008 and T015 use mmtf.fetch, and in my hands cannot be completed, since I have not yet found a workaround. Please note that I am NOT a python expert. Thanks, — Reply to this email directly, view it on GitHub <#63 (comment)>, or unsubscribe <https://github.com/notifications/unsubscribe-auth/AA53AEEDGD7RIVMWWV2MO6L3VGFJ3AVCNFSM6AAAAACH2MH7HOVHI2DSMVQWIX3LMV43OSLTON2WKQ3PNVWWK3TUHMZTGNBYHAYTSOJWGE> . You are receiving this because you are subscribed to this thread.Message ID: ***@***.***>