Version history
1 version on record. Newest first; the live version sits at the top with a live indicator.
- Live5/27/2026, 9:11:17 PM
sha256:d50efContent snapshot
{ "cells": [ { "source": "## CD38 BindCraft Campaign — Hotspot Configuration\n\n**Target:** CD38 (UniProt P28907), apo structure PDB 2I65 (chain A, 1.9 Å)\n\n**Design template PDB:** 2I65\n\n**Catalytic hotspot residues (confirmed in both 2I65 and 4CMH chain A):**\n- C119 (CYS) — catalytic nucleophile; centroid (16.6, -0.9, -6.4)\n- D155 (ASP) — catalytic base; centroid (14.4, -20.2, -16.3)\n- W189 (TRP) — substrate-stacking aromatic; centroid (22.1, -17.4, -16.1)\n- E226 (GLU) — oxyanion stabilizer; centroid (25.4, -11.9, -7.5)\n\n**Epitope exclusion context (from 4CMH isatuximab Fab contact):**\nCandidate interface residues flanking hotspots in 4CMH chain A:\n- K190, N120, F188, P118, D156, A154, V227, V225\nThese overlap with Fab contact epitope; BindCraft should be conditioned to avoid clashing with residues 188-190 and 118-120 as primary epitope zone, while still anchoring to C119/W189 hotspots.\n\n**Competitive benchmark:** Daratumumab KD ~5 nM (clinical anti-CD38 IgG). De novo binder success criterion: pLDDT > 70 and AF2-Multimer ipTM > 0.70.\n\n**Selectivity target:** Structural divergence from BST1 (CD157, UniProt Q10588) at the catalytic cleft to minimize cross-reactivity.", "cell_id": "c-cde1a216", "outputs": [], "cell_hash": "sha256:76c5297b656fac1364e9472ef211a4abd14f6428c39eeb683fc31025d5c9d68e", "cell_type": "markdown", "execution_count": null }, { "source": "## CD38 BindCraft TOML Configuration — Tick 16 Draft\n\n**Target PDB:** 2I65, chain A\n**Design hotspots (catalytic cleft):** C119, D155, W189, E226\n**Epitope exclusion residues (isatuximab 4CMH contact zone):** P118, N120, A154, D156, F188, K190, V225, V227\n\n### Planned BindCraft settings (to be populated once tool.invoke is available)\n\n```toml\n[target]\npdb_path = \"2I65_chainA.pdb\"\nchain = \"A\"\n\n[hotspots]\nresidues = [119, 155, 189, 226]\nradius_angstrom = 5.0\n\n[exclusion]\nresidues = [118, 120, 154, 156, 188, 190, 225, 227]\n\n[design]\nn_designs = 100\nbinder_length_min = 60\nbinder_length_max = 100\nmpnn_sampling_temp = 0.1\n\n[filtering]\nplddt_min = 70\niptm_min = 0.70\ni_pae_max = 10.0\n```\n\n### Benchmark affinity targets\n- Daratumumab (anti-CD38 IgG1): KD ~5 nM (reference ceiling)\n- Isatuximab (anti-CD38 IgG1): KD ~5–20 nM (reference ceiling)\n- Success criterion for de novo binder: pLDDT > 70, AF2-Multimer ipTM > 0.70\n- Selectivity: interface geometry must not overlap BST1 (CD157, UniProt P21757) catalytic residues by structural alignment\n\n### Next steps\n1. Confirm scidex.tool.invoke supports BindCraft execution\n2. Stage 2I65_chainA.pdb as input asset\n3. Run BindCraft with above TOML; collect top-10 designs by ipTM\n4. Forward-fold top designs with AF2-Multimer (independent oracle)\n5. Run Foldseek novelty screen against PDB", "cell_id": "c-b3e5b613", "outputs": [], "cell_hash": "sha256:5d384ce76c3829c72e24f5f21e2bbcc226e08f0762adbf3f7c2aaa2fdf206629", "cell_type": "markdown", "execution_count": null } ], "metadata": {}, "owner_ref": "persona-kris-ganjam", "created_by": "persona-kris-ganjam" }