Team

Confidentiality. This dataset is near submission. All data shared with the team must be kept confidential until publication. Talk to Dale about sharing additional data beyond what’s in the starter pack.

Why this project

PQBP1 is a 265-aa, largely disordered adapter that’s been impossible to pin down by cryo-EM or crystallography. The Chanda lab has assembled NMR, HDX-MS, XL-MS, and biochemistry data on how PQBP1 interacts with oligomeric ligands and recruits cGAS to trigger inflammation — a mechanism implicated in Alzheimer’s and Parkinson’s. The piece that’s still missing is a structural picture, and the hackathon outputs feed directly into a manuscript in preparation.

What a team could build

  • Convert the lab’s NMR, HDX-MS, and XL-MS restraints into HADDOCK constraint files.
  • Run HADDOCK across the three target ligands (HIV-1 hexamer, Tau fibrils, α-syn fibrils) and assess convergence with MDanalysis and TM-comparison.
  • Generate many models without user bias — let the data and the scoring drive ensemble selection.

Minimum viable demo: converged HADDOCK ensembles for the PQBP1 WW domain against at least one oligomeric ligand, plus a short writeup comparing binding surfaces across targets.

Stretch directions

  • Model the tri-molecular complex of PQBP1 (265 aa) + HIV-1 hexamer (221 aa × 6) + human cGAS (522 aa).
  • Coarse-grained simulations of how PQBP1 senses ligands through conformational selectivity.
  • Cross-check HADDOCK ensembles against the lab’s existing Alphalink2 XL-MS model of unbound PQBP1.

Data

Mix of PDB, plain text, and FASTA — primarily structural restraints plus reference structures. The starter pack pairs published WW-domain constraints (Nat. Commun. 2021) with the lab’s NMR, HDX-MS, and XL-MS data, and includes an Alphalink2 model of unbound PQBP1. Preliminary low-confidence cryo-EM maps are available on request.

Reference structures (PQBP1, HIV-1 capsid, Tau, α-syn, cGAS) can be pulled from UniProt on AWS.

Background reading

Skills you’ll build

Integrative structural modelling with HADDOCK; ingesting NMR, HDX-MS, and XL-MS restraints into 3D models; structural-bioinformatics workflows with MDanalysis and TM-comparison; reasoning from sparse experimental data toward biophysical mechanism.