sokrypton / sokrypton/ColabFold

Multimer error on ColabFold 1.6.0

Open
#817 3 comments 0 reactions 0 assignees View on GitHub

Nobody has claimed this yet.

Dominant language
Jupyter Notebook
Stars
2.9k
Forks
747
PR merge metrics
No merged PRs in 30d

Description

Hi,

After upgrading from ColabFold 1.5.5 to 1.6.0, I encountered a TypeError when running in multimer mode. The error does not occur in monomer mode.

Error logs:

2026-03-13 22:30:35,112 Running colabfold 1.6.0
2026-03-13 22:30:41,693 generated new fontManager
2026-03-13 22:30:42,041 Running on GPU
2026-03-13 22:30:42,180 Found 6 citations for tools or databases
2026-03-13 22:30:42,180 Query 2/1: BTB-domain_0 (length 59)
2026-03-13 22:30:42,383 Could not generate input features BTB-domain_0: Calling np.sum(generator) is deprecated.Use np.sum(np.fromiter(generator)) or the python sum builtin instead.
Traceback (most recent call last):
  File "/usr/local/lib/python3.12/site-packages/colabfold/batch.py", line 1469, in run
    = generate_input_feature(query_seqs_unique, query_seqs_cardinality, unpaired_msa, paired_msa,
      ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
  File "/usr/local/lib/python3.12/site-packages/colabfold/batch.py", line 931, in generate_input_feature
    input_feature = process_multimer_features(features_for_chain, min_num_seq=max_seq + 4)
                    ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
  File "/usr/local/lib/python3.12/site-packages/colabfold/batch.py", line 850, in process_multimer_features
    np_example = feature_processing.msa_pairing.merge_chain_features(
                 ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
  File "/usr/local/lib/python3.12/site-packages/alphafold/data/msa_pairing.py", line 426, in merge_chain_features
    np_chains_list = _merge_homomers_dense_msa(np_chains_list)
                     ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
  File "/usr/local/lib/python3.12/site-packages/alphafold/data/msa_pairing.py", line 391, in _merge_homomers_dense_msa
    _merge_features_from_multiple_chains(chains, pair_msa_sequences=True)
  File "/usr/local/lib/python3.12/site-packages/alphafold/data/msa_pairing.py", line 363, in _merge_features_from_multiple_chains
    merged_example[feature_name] = np.sum(x for x in feats).astype(np.int32)
                                   ^^^^^^^^^^^^^^^^^^^^^^^^
  File "/usr/local/lib/python3.12/site-packages/numpy/_core/fromnumeric.py", line 2419, in sum
    raise TypeError(
TypeError: Calling np.sum(generator) is deprecated.Use np.sum(np.fromiter(generator)) or the python sum builtin instead.
2026-03-13 22:30:42,386 Done

Command used:

colabfold_batch --jobname-prefix BTB-domain --save-all  --random-seed 654321 --model-type=alphafold2_multimer_v3 --amber --use-gpu-relax sp_Q9Y2Y4_ZBT32_HUMAN_29-87_Homodimer_Zinc_finger_and_BTB_domain-containing_protein_32_OS_Homo_sapiens_OX_9606_GN_ZBTB32_PE_1_SV_1.a3m predictions

Do you have any ideas on how to solve this problem?

Contributor guide

Open the contributing guide

First steps

  1. Read the whole issue, then the project's contributing guide.
  2. Comment on the issue to say you are picking it up — it saves two people doing the same work.
  3. Fork the repository and make your change on a branch.
  4. Open a pull request that references the issue number.

Research direction

Start with the multimer traceback in colabfold/batch.py, then inspect alphafold/data/msa_pairing.py at _merge_features_from_multiple_chains. Reproduce the failure with the provided colabfold_batch command in multimer mode and compare it with monomer mode. Done means the multimer run no longer raises the NumPy TypeError.

Written by the indexing model from the issue text.

Assessment

Tech stack
numpy, python
Domain
bioinformatics, machine-learning
Issue type
Bug
Difficulty
2/5
Estimated time
1-3 hours
Activity status
Quiet
Clarity
Mostly clear
Newbie friendliness
55/100

Get new issues in your inbox

A short digest of beginner-friendly GitHub issues.