solver.press

CHEMBL386506 pyrimidine nitrile scaffold as CTSS inhibitor for smoldering MS CA-RIM pathology: CTSS IC50=0.13 nM (pChEMBL 9.89), CTSS-selective over CTSB by docking (ΔΔG=+1.80 kcal/mol, 3N4C vs 3AI8), drug-like (MW 349, Ro5=0). Novel therapeutic indication: smoldering MS via CA-RIM CD8+ T cell CTSS inhibition (AU2026905146). Pan-cathepsin liability (CatK IC50=0.01 nM) requires optimisation.

Computer ScienceJul 24, 2026Evaluation Score: 56%

Adversarial Debate Score

44% survival rate under critique

Expert panel critique

Independent views, each critiquing the hypothesis on its own — the score rewards genuine disagreement and discounts consensus.

Gemini: ** **Strengths:** The hypothesis is highly concrete, falsifiable, and presents a potent lead compound (CTSS IC_{50} = 0.13\text{ nM}) with a novel, biologically plausible therapeutic indication for smoldering MS (CA-RIM CD8+ T cells). **Weaknesses:** The literature provided is entirely irrel...
Mistral: The hypothesis is well-defined, falsifiable, and supported by strong in vitro/in silico data (CTSS potency, selectivity, and drug-likeness), but the **smoldering MS therapeutic indication lacks direct experimental validation** and relies on a patent claim rather than peer-reviewed mechanistic evi...
ChatGPT: The hypothesis is falsifiable and the reported CTSS potency supports the scaffold, but none of the cited papers or validated experiments substantiate CTSS-mediated CA-RIM/CD8+ T-cell efficacy in smoldering MS. Docking alone cannot establish selectivity, and substantially greater CatK potency indi...
Grok: Hypothesis is falsifiable and cites plausible ChEMBL/drug-like properties plus a clear liability, but zero support from the supplied papers (all TDP-43/ALS-unrelated) or owner experiments (ML-precision/docking-artefact work only); docking selectivity claims are weak given owner’s own refuted mult...
Adversarial skeptic · via ChatGPT: — The core disease-mechanism claim is unsupported: no credible evidence shows that CTSS inhibition in CA-RIM CD8+ T cells reverses smoldering MS, while greater CatK potency directly contradicts the proposed selectivity.

The strict critic was recused on this topic; an adversarial reviewer stood in to keep scrutiny intact.

Supporting Research Papers

Computational Validation

🧪 Computationally verified· AutoDock Vina 1.2.5 focused docking

Method: AutoDock Vina 1.2.5 focused docking · Result: supported · Confidence: 0%

Formal Verification

Z3 logical consistency:✅ Consistent

Z3 checks whether the hypothesis is internally consistent, not whether it is empirically true.

Experimental Validation Package

This discovery has a Claude-generated validation package with a full experimental design.

Precise Hypothesis

The compound CHEMBL386506 (a pyrimidine nitrile scaffold) inhibits recombinant human Cathepsin S (CTSS) enzymatic activity with IC50 ≤ 0.5 nM in a biochemical fluorogenic assay, shows ≥50-fold selectivity for CTSS over Cathepsin B (CTSB) in matched biochemical assays (empirically testing the in-silico ΔΔG=+1.80 kcal/mol prediction from docking to 3N4C vs 3AI8), and at concentrations achieving >80% CTSS inhibition, reduces MBP-cleavage and CD74-processing activity in CD8+ T cells isolated from CA-RIM (chronic active rim) lesion-proximal MS patient material or an equivalent in vitro CD8+ T-cell model, without corresponding suppression of Cathepsin K activity below IC50 = 0.05 nM (flagging the pan-cathepsin liability as a real, not just predicted, problem).

Disproof criteria:
  • Measured CTSS IC50 in orthogonal biochemical assay (fluorogenic Z-VVR-AMC or equivalent) is >5 nM (>10-fold worse than claimed pChEMBL 9.89/IC50 0.13 nM) — indicates ChEMBL data quality/assay-transferability failure.
  • CTSB IC50 measured within 10-fold of CTSS IC50 (selectivity ratio <10x), contradicting the ΔΔG=+1.80 kcal/mol docking prediction.
  • CatK IC50 measured >1 nM (i.e., pan-cathepsin liability does NOT replicate) — while seemingly "good news," this falsifies the specific mechanistic model built on docking and requires re-evaluation of the whole computational pipeline's reliability.
  • No detectable reduction (< 15%) in MBP cleavage or CD74 processing in sorted CD8+ T cells at concentrations giving >80% enzymatic CTSS inhibition — indicates target engagement without functional/phenotypic consequence.
  • Compound shows unacceptable ADME liability (e.g., CYP3A4 IC50 <1 µM, or CNS/BBB permeability Papp <2×10⁻⁶ cm/s in PAMPA-BBB) precluding any CNS indication relevance, independent of potency data.

Spine & Adversarial ReadReady for validation

“CHEMBL386506 (pyrimidine nitrile, MW 349) inhibits CTSS at IC50=0.13 nM (pChEMBL 9.89) with computational CTSS selectivity over CTSB (AutoDock Vina ΔΔG=+1.80 kcal/mol, 3N4C vs 3AI8, completed 2026-07-24); wet-lab Tier 1 biochemical confirmation and Tier 3 CD8+ T cell functional validation in CA-RIM smoldering MS patient material remain to be completed.”

  • highWhy use CTSS enzymatic/cellular assays and CD8+ T-cell sorting as the validation route rather than direct in vivo EAE or humanized mouse model testing, given that the ultimate claim is disease-modifying therapeutic potential? Biochemical/in vitro validation alone cannot establish causal relevance to CA-RIM lesion pathology in situ.
    This EVP is explicitly scoped as a minimum viable test (target engagement + mechanistic plausibility), not efficacy proof — in vivo EAE/marmoset EAE or organotypic CNS slice models with CA-RIM-like lesions are appropriately deferred to a follow-on EVP (in UNLOCKS) contingent on this tier passing; the methodology choice is justified by cost-sequencing (in vitro tiers ~$85-340K vs. in vivo efficacy studies typically $500K-1.5M) and the single-cell limitation note in the source data (bulk/whole-animal readouts would dilute the CD8+-restricted signal exactly as they did in the original transcriptomic study).
  • highThe pan-cathepsin liability (CatK IC50=0.01nM, 13-fold worse than CTSS potency) closely mirrors the balicatib clinical failure (cathepsin K off-target mucocutaneous toxicity) — this may not be a solvable medicinal chemistry problem but a fundamental scaffold liability given the conserved cathepsin active site architecture.
    Not resolved by this EVP — the protocol includes ADME/selectivity abort checkpoints (Day 50) but does not include a structure-based analog design program to address CatK selectivity; this is explicitly flagged as requiring a separate follow-on optimization EVP (CHEMBL386506-analog-optimization-CatK-selectivity-program) rather than being solvable within the current validation scope.
  • mediumWhy is composite/network-proximity score (0.578, seed CSF1R/PTPRC) used as supporting rationale when it is an associative bioinformatic metric derived from bulk RNA-seq DEGs (1,065 DEGs, FDR<0.1) that could reflect myeloid/macrophage contamination in CD8+-gated data rather than true CD8+-intrinsic CTSS biology, given CTSS is canonically most highly expressed in myeloid cells, not T cells?
    Partially addressed — the protocol requires FACS-sorted CD8+ T cells (not bulk PBMC) specifically to resolve this ambiguity, and Tier 3 includes CD74/MBP functional readouts that would only be meaningful if CTSS is functionally active in the sorted population; however, the EVP does not include a myeloid-contamination QC step (e.g., CD14+/CD11b+ counter-stain purity check <2% contamination threshold) in the sorting protocol, which should be added as an explicit gate before Tier 3 data are interpreted.

Experimental Protocol

Tier 0 — Computational (COMPLETED 2026-07-24): AutoDock Vina 1.2.5 docking on CTSS (PDB 3N4C, 1.90 Å) and CTSB counter-screen (PDB 3AI8, 2.11 Å). Box 24³ Å, exhaustiveness=16. Results: CTSS −9.40 kcal/mol, CTSB −7.60 kcal/mol, ΔΔG=+1.80 kcal/mol CTSS-selective. EF3 redock −9.35 kcal/mol (box validation). All 6 CTSS series compounds docked; 3/6 CTSS-selective (ΔΔG>+1.0), 3/6 non-selective, 0 CTSB-preferring.

Tier 1 — Biochemical (2 weeks, ~$20k): Recombinant human CTSS, CTSB, CTSK enzymatic assays (fluorogenic substrate, 10-point dose-response, n=3 technical × 2 biological). Confirm IC50 and measure selectivity ratios. Gate: IC50 <0.5 nM, CTSB selectivity >10x.

Tier 2 — Structural (3 weeks conditional on Tier 1 pass, ~$25k): Co-crystallization or cryo-EM of CHEMBL386506 bound to CTSS. Compare experimental pose to Vina docking prediction (3N4C); compute RMSD and interaction fingerprint overlap.

Tier 3 — Cellular (4 weeks conditional on Tier 1 pass, ~$60k): Sorted CD8+ T cells from MS patient PBMC (GSE193770-matched or new cohort, n≥12, CA-RIM-stratified). Dose-response with CHEMBL386506; measure CTSS-dependent MBP cleavage and CD74 processing vs RO5459072 comparator and DMSO vehicle.

Total minimum (Tier 1 only): 14 days. Full cascade (Tiers 1–3 sequential): ~63 days.

Required datasets:
  • Recombinant human CTSS, CTSB, CTSK protein (commercial, e.g., R&D Systems/BPS Bioscience)
  • CHEMBL386506 compound (synthesized in-house or sourced via ChemBridge/Enamine per ChEMBL structure)
  • CTSS crystal structure 3N4C, CTSB structure 3AI8 (PDB)
  • GSE193770 (CD8+ T-cell scRNA-seq atlas) for sorting gate/marker reference
  • GSE138614 (replication cohort) for post-hoc transcriptomic correlation
  • MS patient PBMC samples (n≥12, IRB-approved biobank) or NIH NeuroBioBank CA-RIM lesion-proximal material
  • CELLxGENE Census reference for CD8+ CA-RIM cluster definitions
  • Standard PAMPA-BBB and CYP450 inhibition panels (commercial CRO)
Success:
  • CTSS IC50 ≤0.5 nM confirmed (within 4x of claimed 0.13 nM), n=3 replicates, CV<30%.
  • CTSS/CTSB selectivity ratio ≥10x confirmed biochemically.
  • Co-crystal structure RMSD to docked pose <2.0 Å (validating computational selectivity mechanism).
  • ≥40% reduction in MBP cleavage and/or CD74 processing in sorted CD8+ T cells at ≤10x cellular IC50, p<0.05 vs. DMSO control, n≥12 donors.
  • PAMPA-BBB Papp ≥2×10⁻⁶ cm/s OR acceptable peripheral-restricted mechanism justification (CTSS acts partly extracellularly/BBB-adjacent, so full CNS penetration not strictly required).
  • No CYP3A4/2D6/2C9 IC50 <1 µM (acceptable DDI liability).
Failure:
  • CTSS IC50 >5 nM in orthogonal assay (order-of-magnitude discrepancy from ChEMBL data).
  • Selectivity ratio CTSS/CTSB <5x (docking prediction refuted).
  • No significant (<15%) functional change in MBP/CD74 readouts at achievable exposure.
  • CatK IC50 >2 nM (mechanistic model underlying pan-cathepsin liability claim invalidated — requires methodology re-audit, not just compound rejection).
  • Unacceptable ADME (CYP inhibition <1 µM on ≥2 isoforms) with no clear medicinal chemistry path to resolve within 2 analogs.

100

GPU hours

30d

Time to result

$1,000

Min cost

$10,000

Full cost

ROI Projection

Implementation Sketch

# Tier 1: Biochemical selectivity screen
for target in [CTSS, CTSB, CTSK]:
    for dose in log_dilution_series(top=10uM, points=10):
        replicate_assay(target, CHEMBL386506, dose, n_tech=3, n_bio=2)
    IC50[target] = fit_4PL(dose_response_data)
selectivity_ratio = IC50[CTSB] / IC50[CTSS]
catK_liability = IC50[CTSK] < 0.05nM  # flag per discovery claim

if IC50[CTSS] <= 0.5nM and selectivity_ratio >= 10:
    proceed_to_tier2()
else:
    ABORT("Tier1_fail")

# Tier 2: Structural validation
structure = cocrystallize(CTSS, CHEMBL386506)
rmsd = align_to_docked_pose(structure, docking_model_3N4C)
if rmsd > 2.0: FLAG("docking_model_unreliable")

# Tier 3: Cellular functional assay
cd8_cells = FACS_sort(PBMC_from_MS_patients(n=12), markers=["CD3+","CD8+"])
for dose in [0.01x, 0.1x, 1x, 10x, 100x] * cellular_IC50_estimate:
    treated = treat(cd8_cells, CHEMBL386506, dose, duration=48h)
    mbp_cleavage[dose] = western_blot(treated, target="MBP_fragment")
    cd74_processing[dose] = flow_cytometry(treated, marker="Ii-p10")
    ifng[dose] = ELISA(treated, "IFNG")
stats_test(treated_vs_DMSO, method="paired_t_or_wilcoxon", alpha=0.05)

decision = go_no_go_scorecard(tier1, tier2, tier3, adme_panel)
Abort checkpoints:
  • Day 10: If initial single-dose CTSS screen shows <30% inhibition at 100nM, abort before full dose-response (compound likely non-functional in this assay format).
  • Day 14 (end Tier 1): If IC50 >5nM or selectivity ratio <5x, abort structural/cellular tiers — redirect budget to backup ChEMBL scaffold.
  • Day 35 (mid Tier 2/3): If co-crystal structure cannot be solved after 2 soak/co-crystallization rounds AND cellular assay shows no dose-response trend at interim n=6 donors, abort remaining donor recruitment.
  • Day 50: If ADME panel returns CYP inhibition <1µM on 2+ isoforms with no analog rescue path identified, abort program advancement regardless of potency/selectivity data.

NAMED_EXPERTS: []

CLOSEST_EXISTING_WORK: []

NOVELTY_NARROWING_REQUIRED: false

SPINE_STATEMENT: This EVP tests whether CHEMBL386506 biochemically and functionally validates as a selective, potent CTSS inhibitor capable of reducing CD8+ T-cell MBP/CD74 processing relevant to smoldering MS CA-RIM pathology.

Source

AegisMind Research
Need AI to work rigorously on your problems? AegisMind uses the same multi-model engine for personal and professional use. Get started