Radicular: audit the evidence behind a non-sedating peptide analgesic claim

Create a decision-ready evidence audit separating the published support for an α2δ-1 decoy peptide from the evidence still needed for Radicular's proposed analgesic. The purchased result is a traceable claim table and a short, evidence-calibrated replacement research rationale.

The OpenLabs Radicular project proposes drug-like analogs of an α2δ-1 decoy peptide for lumbosacral radiculopathy. This is a literature validation task using the two fixed primary sources below, not a request to design analogs or run experiments.

Assess these six claims as separate rows, with exactly these IDs:

• C1: α2δ-1 physically interacts with NMDA receptors in the biological systems studied by Chen et al.
• C2: The published decoy peptide disrupts that interaction and changes a relevant functional readout.
• C3: The published peptide reduces mechanical hypersensitivity in a nerve-injury animal model.
• C4: Those results establish efficacy in humans with lumbosacral radiculopathy.
• C5: Those results establish that the peptide avoids sedation and dependence liabilities.
• C6: PRECISE establishes a benefit of pregabalin over placebo for sciatica leg pain at week 8.

These are questions to audit, not assumed truths. For each, choose supported, partly supported, contradicted, or not established by the required sources. Define the labels in your report: contradicted requires an opposing observation; not established means the necessary evidence is absent from this bounded source set. An absence finding must not be described as proof that an effect is impossible.

Submit both required files as bytes in the Submission; links to files do not replace them.

1. claim-audit.csv: one row for each C1–C6, with columns claim_id, verdict, source_id, locator, observed_result, population_or_system, comparator_or_control, endpoint, limitation, and reasoning. A locator is a named section and figure/panel, table, or abstract subsection, as applicable. Include study-reported numerical results with units and uncertainty when the cited result supplies them; do not invent missing numbers or digitize plots. If a claim is unsupported, identify the sections checked and the closest relevant result, then explain the missing link.
2. evidence-memo.md: a 600–1000 word memo containing the label definitions, source bibliography with DOI/PMID and access date, a synthesis of the animal-to-human and efficacy-to-safety gaps, a replacement research rationale of at most 150 words, and three ranked questions for documentary due diligence. Explain which missing evidence each question would resolve. These must be information requests, not laboratory procedures or treatment advice.

The table is the evidence ledger; the memo must reference claim IDs rather than introduce uncited new scientific claims. Cover both sources. This is a bounded audit, not a systematic review: optional additional reading does not enlarge the required corpus or alter the six questions. Do not count an abstract and full text of the same study as independent evidence.

Required primary sources, publicly accessible without payment:

• S1: Chen et al., 2018, *The α2δ-1-NMDA Receptor Complex Is Critically Involved in Neuropathic Pain Development and Gabapentin Therapeutic Actions*. DOI 10.1016/j.celrep.2018.02.021, PMID 29490268. PMC full text. The main article and its figure legends are the required evidence boundary; supplementary files are optional and cannot be necessary to pass.
• S2: Mathieson et al., 2017, *Trial of Pregabalin for Acute and Chronic Sciatica*. DOI 10.1056/NEJMoa1614292, PMID 28328324. PubMed abstract, with the same indexed abstract available through Europe PMC public record API. Only the structured abstract is required; publisher access is not required.

The governing versions are the identified 2018 article and 2017 abstract as publicly available on September 22, 2026. Later research does not change this audit. Corrections publicly associated with these records by that date must be disclosed if they affect an extracted result. Source records are authoritative for published observations; neither the project description nor this challenge establishes a therapeutic finding.

Background origin, not additional required input: Radicular project and linked June 10, 2026 post. The six claims printed above fix the scope if the project text changes.

Only Submissions meeting every Acceptance Criterion are eligible. Rank eligible Submissions first by the strength of the link between each verdict and its cited observation: precise locators, correct context, and explicit reasoning take priority over prose style. Next prefer a replacement rationale that makes fewer unsupported extrapolations while retaining the actual research question. Next prefer due-diligence questions whose answers would resolve the most consequential evidence gaps, with the reason stated in the memo. If still tied, the lowest numeric on-chain Submission ID wins. If only one Submission qualifies, it wins; if none qualifies, the outcome is no_valid_submission.

Missing a required file or claim row, invented evidence, material misrepresentation as defined above, or inclusion of prohibited material under Out Of Scope makes a Submission ineligible when the failure is established. A negative finding by itself is not disqualifying.

New wet-lab work, peptide sequences or analog design, organism engineering, experimental protocols, patient-specific advice, therapeutic recommendations, and claims that this literature audit proves clinical safety or efficacy.