Follow the claim
to the evidence.

03 / External cohort recheck

The main result survives the flagged-array check.

Butler et al. excluded one low-quality pooled array; our v0.9 package already included that exclusion as a sensitivity analysis. A separate independent-preparation array is highlighted in their Figure 7. Removing it is an additional post hoc sensitivity, not a second author-mandated exclusion.

The skin-and-blood comparison is secondary and more sensitive to this array. PhenoAge and DNAmTL retain lower disagreement. Horvath 2013 and Hannum remain outside the eligible primary external comparisons because of clock-probe coverage; no new superiority test is claimed.

We reanalysed the supplied v0.9 score table, not raw methylation intensities. SD is the square root of summed within-donor squared deviations divided by the sum of donor-specific residual degrees of freedom. The download contains all four exclusion scenarios and both preparation arms. Shared slides and only four donors limit generalization.

Butler et al., methods and Figure 7 · Code, inputs and complete results

04 / A proposed biological next step

Can coordinated function recover?

What measurements would distinguish recovery of collective cellular organization from a change in molecular biomarkers alone?

This is a study-design brief for scientific discussion, not a ready-to-run protocol, funded commitment or completed experiment. The experimental model and intervention must be selected with appropriate biological expertise.

Before including a methylation clock, establish its coverage and suitability for the chosen tissue and culture state. Blood-array duplicate error cannot be carried into this experiment as its uncertainty estimate.

Choose the readout before the intervention

Airway-barrier recovery is a candidate, not a validated winner. Compare it with fibroblast wound closure (growth and migration must be separated) and muscle-force recovery (specialized model and readout required). Choose with the experimental team based on feasibility, donor variation and the mechanism being tested.

For an airway model, a candidate primary endpoint is tracer-permeability recovery at a prespecified time, supported by morphology and viability. Electrical resistance (TEER) is useful but depends on ion transport as well as junction integrity; a bioelectric intervention could change TEER without repairing the barrier. Normalize resistance for area and blank inserts, standardize measurement conditions, and include an orthogonal permeability readout. TEER is not a measurement of individual-cell membrane potential.

Primary airway-model methods and ion-transport measurements support the assay distinction; the proposed endpoint hierarchy is Karam’s design recommendation.

Donor-level precision, with explicit assumptions

Let τ be the SD of true donor-specific treatment contrasts, σw the independent insert residual SD, and k the inserts per arm. Then Var(D) = τ² + 2σw²/k. This assumes equal residual variance and independent arm residuals; estimate the variance of donor contrasts directly if those assumptions fail. Inserts do not create additional donors.

Illustration only: Δ = 10, τ = 10 and σw = 8 in the endpoint’s own units, with k = 3, gives contrast SD 11.94 and 14 evaluable donors for 80% power at two-sided α = 0.05, using noncentral-t power. These are assumed inputs, not pilot measurements. Allow separately for attrition; confirm feasibility and uncertainty in a pilot.

Composition change can be part of repair. If it accompanies improved function, report both and leave the mechanism unresolved; do not automatically adjust away a post-treatment mediator. A failed mechanism-blockade experiment is interpretable only with demonstrated target engagement and appropriate viability controls.

Download the sensitivity analysis, corrected planning code and review record

06 / Scientific context & attribution

Build on the work. Keep the credit.

Michael Levin and collaborators study developmental bioelectricity, morphogenesis and the capabilities of cellular collectives. Their work motivates questions about coordination across scales. Karam’s contribution presented here is a measurement-focused computational program and a proposed experimental direction.

• Levin Lab: research overview — primary description of the lab’s questions and approach.
• Karam’s bioelectricity source register — named authors, study contexts and access limitations.
• ClockQC methods and provenance — public measurements, computational comparisons and reproduction scope.

External papers remain their authors’ work. No partnership, endorsement or shared intellectual property with Tufts or the Levin Lab is implied.

Claim-by-claim evidence

What “memory” would have to mean here

Specify an observable state, a controlled history, a retention interval and a future response. Compare history-dependent predictions against simpler explanations such as continuing stimulation, composition change and passive mechanics. A stable shape or correlated biomarker alone does not identify the storage mechanism.

Reading and verification scope

The 2025 Anthrobot and Butler papers were read through their introductions, results, discussions and main methods, including figure captions. Anthrobot S5 was inspected separately. Relevant airway assay methods were read; that paper’s entire body was not audited. The PC-clock paper remains contextual prior work. Its main text and methods were read from the public author manuscript for this cleanup; its supplementary analyses and raw data were not independently reproduced in this pass.

This is not a completed 24-paper full-text review or an independent audit of every linked venture. The project descriptions in the short briefing are checked against their own current public pages; this establishes what they report, not that their results are independently validated.

• Gumuskaya et al., 2025 — main article and S5 read as specified above; raw data not obtained.
• Butler et al., 2025 — main article read; Figure 7 visually checked; relevant author scripts inspected. Karam’s score SD estimator differs from the paper’s per-CpG SD-ratio estimator.
• Higgins-Chen et al., 2022 — foundational PC-clock method; main text and methods read in the public author manuscript. No new experimental result is attributed to this review.
• Primary human airway methods — assay sections read; study selection remains a proposal.

What still needs a different environment or evidence

Fresh raw-array processing, sample-level Anthrobot data, a completed broader literature review, and host-side Apache/PHP and email-delivery tests remain open. Mobile-width checks in a browser are not physical-device testing. No live deployment or contact with external researchers is claimed.

Sharper methods for the proposed study

Read the endpoint rationale, corrected power illustration and limits of the physical-memory analogy in the 24 October methods note.