Conversation
There was a problem hiding this comment.
💡 Codex Review
Here are some automated review suggestions for this pull request.
Reviewed commit: ffaaadc975
ℹ️ About Codex in GitHub
Your team has set up Codex to review pull requests in this repo. Reviews are triggered when you
- Open a pull request for review
- Mark a draft as ready
- Comment "@codex review".
If Codex has suggestions, it will comment; otherwise it will react with 👍.
Codex can also answer questions or update the PR. Try commenting "@codex address that feedback".
.gitmodules
Outdated
| path = alphafold3 | ||
| url = https://github.com/google-deepmind/alphafold3.git | ||
| branch = main | ||
| url = git@github.com:KosinskiLab/alphafold3.git |
There was a problem hiding this comment.
Keep alphafold3 submodule URL clonable without SSH keys
Switching alphafold3 to git@github.com:KosinskiLab/alphafold3.git makes fresh clones and CI jobs fail in environments that only allow anonymous/HTTPS fetches (the common case for public repos and many runners). This regresses setup reliability because git submodule update --init --recursive now requires preconfigured SSH credentials just for this one submodule.
Useful? React with 👍 / 👎.
| ) -> list[int] | None: | ||
| residue_ids = getattr(chain, "residue_ids", None) | ||
| if residue_ids is not None: | ||
| return [int(residue_id) for residue_id in residue_ids] |
There was a problem hiding this comment.
Handle non-integer residue IDs when slicing AF3 JSON ranges
The new range-slicing path assumes every chain.residue_ids entry is castable to int, but discontinuous/gapped AF3 inputs can legitimately carry insertion-coded IDs like "2A" (which this same change now produces in JSON side data). In that case _chain_residue_ids raises ValueError, so valid JSON feature inputs with ranges fail during prepare_input instead of being sliced.
Useful? React with 👍 / 👎.
No description provided.