Score 15-PGDH binder designs using the ipSAE metric on Lyceum.
npx skills add https://github.com/alex-hh/in-silico-pgdh --skill pgdh-ipsae
Please help me install this Agent Skill. Skill: pgdh_ipsae Source: https://github.com/alex-hh/in-silico-pgdh/tree/main/.claude/skills/pgdh_ipsae Command: npx skills add https://github.com/alex-hh/in-silico-pgdh --skill pgdh-ipsae
This Skill automates the scoring and ranking of protein binder designs for the 15-PGDH target, streamlining the evaluation process and identifying top candidates for further development.
Use the pgdh_ipsae skill to score the binder design located at 'pgdh_campaign/out/boltz2/candidate_1/model_0.cif' with its corresponding PAE file 'pgdh_campaign/out/boltz2/candidate_1/pae.json'.
High-intent search queries and answers about installing and using this skill.
You can score protein binder designs for the 15-PGDH target using the ipSAE metric on Lyceum. This Skill evaluates protein-protein interaction confidence by processing predicted complex structures and PAE files to calculate interaction scores.
The ipSAE metric is used for ranking protein binder candidates by evaluating interaction confidence. It works alongside pDockQ and LIS metrics to filter designs before computationally intensive steps like cross-validation.
To evaluate RFdiffusion or BoltzGen binder predictions against PDB 2GDZ, provide the predicted complex structure file and its corresponding PAE file. The Skill calculates ipSAE and other metrics to rank the designs.
Yes, you can filter protein binder designs before cross-validation using ipSAE scores. The Skill ranks candidates based on ipSAE, pDockQ, and LIS metrics to identify top candidates for further development and submission.
Yes, you need predicted complex structures and PAE files to score 15-PGDH binders. These inputs are required for the ipSAE evaluation process to assess binding potential and rank the designs accordingly.
Using ipSAE for protein design scoring requires precomputed predicted complex structures and PAE files. It applies specifically to the 15-PGDH target and serves as an initial filter before cross-validation rather than a final validation method.
junior1pRank protein binder designs using ipSAE_min, LIS, and pDockQ scores.
alex-hhOrchestrate an end-to-end protein binder design pipeline for 15-PGDH.
alex-hhScore predicted protein-protein interactions using ipSAE, pDockQ, pDockQ2, and LIS metrics.
alex-hhCross-validate 15-PGDH binder designs using Boltz-2 structure prediction.
