nChroma Bio

Phase A — Understand the business

Lens 1 · Company Overview

nChroma Bio is a private, clinical-stage genetic-medicines company in Boston (201 Brookline Avenue) built to do one thing the genome-editing field has spent a decade struggling with: turn a disease gene off durably, without cutting the DNA, and actually get the machine to the right tissue. It was formed in December 2024 by merging two venture-backed startups that were each missing the other's half — Chroma Medicine (the epigenetic "cargo" — editors that write repressive marks onto DNA) and Nvelop Therapeutics (the "delivery" — programmable non-viral vehicles).

What the product is. A single-administration epigenetic silencer. The lead asset, CRMA-1001, is a catalytically-inactive Cas9 (dCas9) fused to proprietary epigenetic effector domains that deposit targeted DNA methylation at hepatitis B viral loci — silencing gene expression without double-strand breaks, and hitting both the episomal cccDNA reservoir and integrated HBV DNA. The pitch is a functional cure for chronic hepatitis B (and ultimately HBV/HDV coinfection) from one dose, where the standard of care is lifelong nucleos(t)ide suppression that cures ~1% of patients.

Business model. Pre-revenue, pre-product, single-clinical-asset biotech. There is no P&L — value is the risk-adjusted NPV of CRMA-1001 plus an option on the delivery platform (LNP for liver, engineered virus-like particles / eVLPs for extrahepatic targets). Monetization is the standard genetic-medicine path: dose-escalate → proof-of-concept → partner or sell to large pharma, or IPO into a data catalyst. nChroma is not selling anything; it is spending venture capital to de-risk a binary scientific bet.

Customers / suppliers / competitors (named).

• End "customer" (commercial, years out): the ~254 million people living with chronic HBV globally, gated through specialist hepatologists and, realistically, a large-pharma commercial partner.
• Key suppliers: contract development & manufacturing orgs (CDMOs) for the dCas9 payload and LNP formulation — unnamed publicly; this is a chokepoint (see Lens 2).
• Competitors: Tune Therapeutics (TUNE-401, the direct epigenetic-silencer rival, ahead in the clinic), plus the broader HBV functional-cure field — GSK (bepirovirsen, ASO, Phase 3), J&J/Janssen (JNJ-3989, siRNA), Vir Biotechnology (VIR-2218, siRNA).

Contract structure. None that bears on revenue — no take-or-pay, no recurring contracts. The only "contracts" are the venture financing terms and (unconfirmed) any platform-delivery collaborations. Capital structure, not commercial structure, is what matters at this stage.

Lens 2 · Supply Chain

For a pre-clinical-stage drug the "supply chain" is the discovery-to-dose value chain, and naming the actual links exposes where nChroma is and isn't differentiated:

Academic IP (Broad / MIT / Whitehead — CRISPRoff, dCas9-effector domains) → Chroma/nChroma in-house editor engineering → dCas9 + effector-domain payload (mRNA/protein) → delivery formulation: LNP (liver) or eVLP (extrahepatic) → CDMO manufacturing (GMP payload + particle fill-finish) → Phase 1/2 clinical sites (Hong Kong, New Zealand, UK) → chronic-HBV patient.

Chokepoints / single-source dependencies (the analyst's job here is to name them):

1. Delivery is the chokepoint the merger was supposed to solve. Chroma had best-in-class editors and no proprietary in-vivo delivery; Nvelop had delivery and no clinical cargo. The entire strategic logic ("a cargo and delivery company over dinner," per CEO Jeff Walsh) is that delivery — not editing — is the rate-limiter in genetic medicine, plagued by "dose-limiting toxicities and delivery inefficiencies." For CRMA-1001 specifically, delivery defaults to LNP to the liver — a commoditized, crowded route (the same LNP-to-hepatocyte path used by every siRNA and many editors), so the proprietary eVLP delivery moat does not protect the lead asset; it protects the next programs.
2. GMP manufacturing of a dCas9-scale payload — large protein/mRNA, CDMO-dependent, no disclosed in-house GMP. A single CDMO slip delays the only value-driving catalyst.
3. cccDNA + integrated-DNA dual targeting — scientifically the differentiator, but it raises the specificity bar (off-target methylation across the genome is the manufacturing-of-confidence chokepoint regulators will probe).

Names or it didn't happen: the upstream IP traces to the labs of David Liu, Keith Joung, Jonathan Weissman, Luke Gilbert, Luigi Naldini, Angelo Lombardo (the six academic founders of Chroma). The downstream clinical chain runs through regulators in Hong Kong (first CTA), New Zealand, and the UK — pointedly not the FDA first (see Lens 10/13).

Lens 3 · Competitive Advantages (moats)

The modality moat (shared, not proprietary). Epigenetic silencing as a class has a genuine advantage over both classical CRISPR and RNA approaches: it modulates expression without double-strand breaks (no genomic-rearrangement / DNA-repair-pathway risk that dogs nuclease editing) yet aims for durability that transient RNA approaches (siRNA, ASO — which require chronic re-dosing) cannot match. That is a real moat for the modality — but nChroma shares it with Tune, Epic Bio, Omega, Moonwalk.

Company-specific moats — and where they're thin:

• IP / founder pedigree (strong on paper). Chroma was built on CRISPRoff and a stack of foundational epigenetic-editing patents from arguably the field's best lab cluster (David Liu — Beam/Prime founder; Keith Joung; Weissman). This is the deepest defensible asset.
• Dual cargo + delivery platform (the merger thesis). Owning both the editor and a non-viral delivery system is rarer than owning either alone — if the eVLP extrahepatic delivery works, it is a platform other editor companies would need.
• Switching costs / network effects: none. This is a drug, not a platform business. There is no installed base, no lock-in. The "moat" is patents + clinical lead, both contestable.

Bargaining power. Essentially none today — pre-revenue, cash-consuming, one asset, dependent on CDMOs and capital markets. Bargaining power is entirely prospective: it accrues only if CRMA-1001 prints clean human data, at which point the leverage flips (pharma needs the asset more than nChroma needs any single partner). Right now the company is the price-taker on capital and manufacturing.

Verdict on the moat: the durable moat is IP + founding science + the dual platform; the lead asset's moat is weak (commoditized LNP-to-liver delivery, a faster rival in the exact same indication). The platform is the bull case; the lead asset is the de-risking event.

Lens 4 · Segments

No revenue → no commercial segments. `` The operating-company segment lens does not apply; the meaningful "segmentation" of a single-asset clinical-stage private is by program and by clinical geography:

By program (the real "segment" mix — 100% concentrated):

By geography (clinical-site footprint, not revenue): first-in-human sites are Hong Kong (first CTA), New Zealand, United Kingdom — deliberately non-US. The geographic "trend" that matters: a non-FDA-first footprint, which lowers cost/speed but raises the question of when (and on what data) a US IND follows (see Lens 13, Q13).

Trend / cause: the value mix has concentrated, not diversified — early Chroma pitched a broad multi-indication platform; capital pressure narrowed nChroma to one liver-led asset to data. Acceleration is binary on the late-2026 readout, not gradual.

---

Phase B — Measure performance (+private + clinical overlay: funding trajectory, pipeline-by-phase, catalyst calendar, traction)

Lens 5 → Funding & Valuation Trajectory (+private swap) + Pipeline-by-Phase (+clinical swap)

Round history (all ``, unaudited):

Cumulative capital into the two franchises ≈ $435M. The honest read: ~$360M had been spent or committed across Chroma + Nvelop before the merger, and the combined company chose to raise only $75M of fresh money — a down-sized reset, not an up-round.

The merger is a capital-allocation tell, not a triumph. Both companies cut staff (Chroma laid off in May 2024; the combined entity took "some reduction in the combined headcount" as "a necessary step"). Two well-funded startups fusing and shedding headcount in a hard biotech tape, raising a modest $75M, is the signature of capital scarcity forcing consolidation — the strong version of "1+1 = a better-funded shot," the weak version of "neither could raise a clean up-round alone." Reality is between, tilted toward necessity.

Pipeline by phase (+clinical — the asset table is the company):

Preclinical efficacy for CRMA-1001 (the data the whole bet rests on): in transgenic and AAV-HBV mouse models, >3-log reductions in HBsAg with durability extending 6 months; at the highest dose up to 90% of treated mice had undetectable HBsAg AND HBV DNA; non-human-primate studies showed sustained silencing >1 year post-dose. These are genuinely strong preclinical numbers — the question (Lens 12/13) is whether they translate to human cccDNA.

Lens 6 · Founder / Management Voice (sentiment) (+private: interviews, not earnings calls)

No earnings calls exist. The available "voice" is founder/exec commentary:

• Jeff Walsh (CEO) frames the merger candidly as opportunistic and necessity-driven — the cargo-meets-delivery "dinner" story, plus explicit acknowledgment that headcount reduction was "a necessary step." Tone: pragmatic, operationally sober — not the promotional hype that should worry a forensic reader. That candor is mildly reassuring.
• The consistent message across 2024→2026 is disciplined and on-strategy: "drive the lead program to meaningful clinical data" and "build a robust pipeline of hepatic and extrahepatic targeted therapies." They delivered on the stated 2025 CTA / 2026 dosing timeline — the CTA authorized Dec 2025, first patient Jan 2026, on schedule. Execution matches narrative, which is the single best signal you can get from a private with no financials.

Sentiment trend: stable-to-constructive, low promotional quotient, milestones hit on time. The "things they stopped saying" tell: early-Chroma talked up a broad multi-indication platform; nChroma narrowed to HBV-first, liver-led, one asset to data — a healthy focusing, but also an admission that capital forces a single shot on goal.

Lens 7 → Cap Table & Syndicate Quality (+private swap) + Mechanism Comps (+clinical swap)

Syndicate quality (the IPO-proximity read): unusually deep and crossover-heavy for a private at this stage. Tier-1 venture (ARCH, Atlas, Newpath, 5AM, Sofinnova, F-Prime, GV, DCVC Bio, Menlo, Omega, Casdin) plus dedicated crossover / public-markets funds: Cormorant (lead), T. Rowe Price, Wellington, Janus Henderson, Sixth Street, Mubadala, Alexandria. A Cormorant-led round with T. Rowe, Wellington and Janus Henderson on the line is a textbook IPO-syndicate-in-waiting — these are the names that mark a company and later anchor a crossover/IPO. But crossover presence ≠ imminence: they came in via the merger reset, the company is single-asset and pre-PoC, so the crossovers are a quality signal and an option, not a near-term S-1 trigger.

Valuation / secondary marks: n/a — private, not disclosed. PitchBook lists a profile but gates the figure behind a paywall; no public post-money or secondary mark is sourceable. Do not infer a number.

Mechanism comps (by target/modality, not P/E — this is the right comp axis for a clinical asset):

The comp that matters is Tune, and it is unambiguously ahead with human data.

Lens 8 · Catalysts / Events that Moved the Story (+private: funding & data events, not stock moves)

No public stock → "what moves the story" = milestones (the proxy for future stock-price catalysts once tradeable):

• Oct 2021 — Chroma $125M launch: established the franchise.
• early 2023 — $135M Series B: kept it alive through the biotech downturn.
• Dec 2024 — merger + $75M: the survival/consolidation event; the down-sized reset.
• Dec 2025 — CTA authorized (HK, then NZ, UK): de-risked regulatory path; on-time execution.
• 28 Jan 2026 — first patient dosed: the asset is officially in humans.
• Late 2026 (pending) — first human clinical data: the catalyst that defines the company. This single readout is worth more than the prior five events combined.

Pattern: the story reacts to (1) cash (can it survive) and (2) clinical de-risking (does the science work in humans). For a future IPO, the late-2026 PoC readout is the make-or-break print.

---

Phase C — Judge people & books (+clinical: + science & exclusivity)

Lens 9 · Management

• Jeff Walsh — CEO (ex-Nvelop CEO). Operator who carried the delivery franchise and now runs the merged company; candid, execution-focused, hit the CTA/dosing timeline on schedule. Less of a public track record than the chairman, but the on-time delivery is the proof.
• Jeff Marrazzo — Chairman (co-founder & CEO of Spark Therapeutics, 2013–2022). This is the strongest single line on the cap table. Marrazzo built Spark from launch to the first FDA-approved gene therapy for a genetic disease (Luxturna), raised ~$1B, struck Pfizer/Novartis partnerships, and sold Spark to Roche for ~$4.8B (Feb 2019) — a ~100× value increase over six years. Having the man who ran the canonical gene-therapy build-and-exit as chairman is a powerful M&A-optionality signal: he knows exactly what a large-pharma acquirer wants to see and how to package an asset for sale.
• Melissa Bonner, PhD — CSO. Leads the science; ex-Chroma.
• Academic founders (Chroma): David Liu (Broad; founder of Beam and Prime — the field's most successful editing-company builder), Keith Joung, Jonathan Weissman, Luke Gilbert, Luigi Naldini, Angelo Lombardo.
• Catherine Stehman-Breen (ex-Chroma CEO) stepped to an advisory role at the merger.

Capital-allocation history: mixed-to-pragmatic. The decision to merge and downsize rather than burn each company's cash chasing parallel programs was disciplined capital allocation under duress — but it is also an admission that prior capital (~$360M) bought less progress than hoped, and that neither standalone could raise a clean up-round. Insider ownership: not disclosed (private); founders/VCs hold, no figures sourceable. Skin in the game: founders are repeat company-builders with reputational capital staked.

Red flags (management): none of the classic forensic ones (no related-party deals, no promotional behavior, no excessive-comp disclosure — because nothing is disclosed). The honest yellow flag is strategic: serial financings and a restructuring-merger reveal a franchise that has consumed a lot of capital for a still-binary, pre-PoC asset.

Archetype: founder-scientist origin (Liu et al.) + professional-operator execution (Walsh/Marrazzo). For a clinical-stage company chasing a pharma exit, that is close to the ideal pairing — science credibility to make the asset, dealmaker credibility to sell it.

Lens 10 · Forensic Red Flags + Science & Exclusivity (+clinical) + Regulatory Findings (required)

Forensic accounting: n/a — private, unaudited, no financial statements. There is no income statement, balance sheet, or cash-flow statement to forensically examine; no revenue recognition, SBC, goodwill, or receivables to interrogate. This is not a clean bill of health — it is the absence of disclosure, which is itself a risk: an outside investor is trusting the syndicate's diligence, not audited numbers.

Science & exclusivity (the +clinical re-point — this is where forensic scrutiny lands for a pre-revenue name):

1. Trial-design / endpoint integrity. The Phase 1/2 (NCT07200193) tests single- and multiple-ascending doses with safety/PK/PD/efficacy endpoints. The real scrutiny: is the efficacy endpoint durable HBsAg loss off-treatment (the true functional-cure bar), or merely on-treatment biomarker repression? Tune set the benchmark at durable repression to 17 months single-dose — nChroma must clear a similar durability bar, not just a transient knockdown.
2. The durability question is the modality's open scientific risk. Peer-reviewed work shows epigenetic HBV silencing has, in some models, been reversed — induced DNA methylation removed (TET-mediated), and treated animals returned to HBV titers comparable to untreated controls. nChroma's own preclinical durability (6 mo mouse, >1 yr NHP) is encouraging but not proof in human chronic infection. This is the single most important forensic line in the dossier: the entire value proposition is "durable from one dose," and durability is exactly what the literature flags as unproven.
3. IP estate / exclusivity: foundational CRISPRoff + epigenetic-editor patents from the Liu/Joung/Weissman cluster — likely strong, but the field is litigation-prone (cf. the CRISPR patent wars) and Tune holds its own TEMPO estate. Freedom-to-operate is a real, unquantified risk. n/a — patent specifics not sourced.
4. Reimbursement/payer path: premature (pre-PoC), but a one-time functional cure for a chronic disease is a favorable payer story if it works.

Regulatory Findings (required sub-section):

• SEC enforcement (EDGAR EFTS — LR + AAER): none possible / none found. nChroma Bio has no CIK — it is private and not required to file with the SEC, so no EDGAR enforcement search applies.
• Non-SEC enforcement (web search — FTC/DOJ/FDA/consent-decree/settlement/penalty): no material findings. No FDA warning letters, DOJ/FTC actions, consent decrees, or penalties surfaced for nChroma Bio or Chroma Medicine across the searches run. The company is pre-commercial with no marketed product, so the surface for enforcement is minimal.
• Item 3 (Legal Proceedings): n/a — no 10-K exists (private, no EDGAR).
• Conclusion: No material regulatory or legal findings — verified via SEC EDGAR EFTS (LR, AAER), web search (FTC/DOJ/FDA/consent-decree/penalty), and the absence of any SEC filing obligation, as of 2026-06-30. Findings are web-only and the company is unaudited per public sources.

---

Phase D — Project & stress-test (+clinical: rNPV + runway-to-catalyst, not EPS)

Lens 11 → rNPV & Runway-to-Catalyst (+clinical + private swap — no EPS line)

There is no EPS to project (pre-revenue) and forecast.ts create is skipped per --watchlist rules. The two questions that actually matter for a private clinical-stage name:

(A) rNPV of the lead asset, CRMA-1001 (illustrative, every input ``):

• Addressable population: ~254M chronic HBV globally; a functional-cure therapeutic realistically reaches a treated, reimbursed subset — assume a peak-treated cohort in the low single-digit millions over time.
• Peak-sales scenario (if approved): a one-time functional cure priced at gene-therapy economics could be a multi-billion-dollar peak franchise — credibly $2–5B peak in a bull approval case.
• Probability of success: a pre-PoC, Phase 1 epigenetic asset in HBV carries a low single-to-low-double-digit PoS — call it ~10–15% to approval.
• rNPV: peak $2–5B × 10–15% PoS × discount for time/cost → a risk-adjusted asset value in the few-hundred-million to ~$1B range, which is consistent with (and a sanity check on) the magnitude of capital already invested ($360M). The honest output: n/a as a precise figure — not sourced; the rNPV is dominated entirely by the binary PoS, which the late-2026 readout will swing by an order of magnitude.

(B) Runway-to-catalyst — the question that actually decides survival:

• Fresh cash: $75M (Dec 2024) + undisclosed cash carried from both pre-merger balance sheets ("multiple years of runway" per the company).
• The next value-inflection catalyst — first human PoC data — is anticipated late 2026.
• Does runway reach the catalyst? Almost certainly yes — the company explicitly guided "multiple years of runway" from Dec 2024, and the catalyst is ~24 months out. The real runway question is whether it reaches the catalyst with enough cushion to raise the NEXT (much larger) round on the data, or whether it hits the readout cash-thin and is forced to raise/partner from weakness. Given only $75M fresh + a single asset, a disappointing readout likely means a distressed raise or a fire-sale partnership.

Brier forecast (the binary that matters, logged conceptually — not via forecast.ts per --watchlist): CRMA-1001 demonstrates durable (off-treatment) HBsAg reduction in the Phase 1/2 readout by end-2027 — p ≈ 0.30.

Lens 12 · Bull vs Bear

Bull case. Epigenetic silencing is the most credible path to a one-dose functional cure for a 254M-patient chronic disease — and Tune's May-2026 human data just de-risked the entire modality (100% biomarker repression at dose levels 2–4, durable to 17 months). That is a gift to nChroma: it proves the mechanism works in humans before nChroma's own readout, converting nChroma from "will the modality work?" to "is nChroma's editor as good as Tune's?" — a far better question to be asking. nChroma brings (a) arguably superior IP/founding science (the Liu/Joung/Weissman cluster, dual cccDNA + integrated-DNA targeting), (b) a chairman (Marrazzo) who has run the exact gene-therapy build-to-$4.8B-exit playbook, (c) a crossover-heavy syndicate (Cormorant, T. Rowe, Wellington) primed to fund an IPO, and (d) a second act — the eVLP extrahepatic delivery platform — that is an option no pure HBV play has. A clean late-2026 readout makes nChroma a prime M&A target or a 2027 IPO candidate.

Bear case (2–3 permanent-impairment risks).

1. Durability fails in humans. The literature shows epigenetic HBV silencing has been reversed in models. If CRMA-1001's methylation doesn't hold off-treatment in chronically infected humans, the one-dose-cure thesis collapses — and takes the rNPV to near-zero. This is the modality's existential risk and it is unproven, Tune's signal notwithstanding (different editor, different effector).
2. Second-place economics. Tune is ~12 months ahead in the identical indication with identical mechanism. In a functional-cure market, first-to-credible-data captures the partnership/pricing high ground; a year-late "me-too epigenetic silencer" may find the differentiated capital and the best pharma partner already taken. Being #2 in the same lane is a structural value haircut.
3. Capital fragility. Single clinical asset, only $75M fresh, on a merger-of-necessity balance sheet. A soft readout, a CDMO delay, or a 2027 biotech-funding winter forces a down-round or fire-sale — permanent dilution/impairment for existing holders.

Pre-mortem (18 months out, thesis broke): the late-2026 readout showed strong on-treatment HBsAg knockdown but the effect waned after dosing stopped — exactly the reversal the literature warned of. Simultaneously Tune's larger Phase 2 confirmed durable cure, locking up the marquee pharma partner. nChroma, cash-thin, raised a punishing down-round and pivoted to the eVLP platform as the surviving story.

Are the multiples too high? No public multiple exists. The implied private value is dominated by the binary readout; "too high / too low" is unknowable until the data prints. n/a.

Contrarian view (what the market refuses to see): consensus frames nChroma as the also-ran to Tune. The contrarian read is that Tune's success is nChroma's call option — modality validation is the hardest gate in biotech, and Tune just paid for it. If nChroma's dual-targeting editor matches Tune's durability, a ~12-month lag is trivial in a market this large and this underserved, and nChroma's superior delivery platform + Marrazzo-grade dealmaking could make it the better-packaged acquisition. Second place in a validated, enormous market is a far better place to stand than first place in an unproven one.

Lens 13 · Devil's Advocate (short-seller)

Dismantling the bull case:

• Revenue concentration = 100% on one unproven asset. There is no revenue and no second clinical program — CRMA-1001 is the company. One readout is binary, total, and ~18 months away. There is no diversification to cushion a miss.
• The moat is weaker than bulls think. The lead asset rides commoditized LNP-to-liver delivery — the proprietary eVLP moat doesn't even touch CRMA-1001. The IP is contestable (CRISPR-style patent wars; Tune's competing estate). "Best founding science" doesn't win clinics — data does, and a competitor has the data.
• The most dangerous competitor bulls underestimate is not Tune — it's GSK. Bepirovirsen is Phase 3 with a 19% functional cure and FDA Fast Track, accepted for review in China. If a re-dosable ASO delivers a "good-enough" functional cure and reaches market years before any editor, the editors' "one-dose" premium has to clear a much higher bar to matter — and editors carry genome-wide off-target-methylation safety questions an ASO doesn't.
• Worst capital-allocation reality: ~$360M consumed across two franchises that then had to merge and lay off staff and could only raise $75M fresh. That is the financial fingerprint of programs that underdelivered relative to capital — a short-seller's "value-destruction-by-consolidation" tell.
• Assumptions that must hold for today's (private) value: (1) human durability holds off-treatment; (2) the late-2026 readout is clean and timely; (3) the company raises its next round before cash pressure bites; (4) freedom-to-operate survives the patent thicket. If clinical efficacy disappoints by 20–30% (e.g. HBsAg loss is real but not durable, or response rates trail Tune), the rNPV doesn't fall 20–30% — it falls off a cliff, because the whole thesis is the durable one-dose cure, and a partial/transient effect reprices the asset toward zero.
• Single scenario that permanently impairs the business: durability reversal in the human readout. Plausibility: moderate — the literature explicitly documents it, and human chronic HBV is a harsher test than transgenic mice. This is not a tail risk; it is the risk.

Lens 14 · Management Questions (ordered by information value)

1. In the Phase 1/2 readout, what is the off-treatment durability endpoint and timepoint — i.e. how long after a single dose must HBsAg suppression hold to count as a functional cure, and how does that compare to Tune's 17-month single-dose data?
2. The literature shows epigenetic HBV silencing can be reversed by host TET demethylation. What in CRMA-1001's effector design prevents re-activation of the methylated loci over months-to-years in chronically infected humans?
3. How much cash runway do you have past the late-2026 readout, and is it enough to fund the next round/partnership from a position of strength rather than necessity?
4. CRMA-1001 uses LNP-to-liver delivery — a crowded, commoditized route. What is your specific differentiation vs. Tune's LNP/TEMPO, and does the proprietary eVLP platform touch the lead asset at all or only future programs?
5. You target both cccDNA and integrated HBV DNA. What is the off-target methylation profile genome-wide, and how will you convince regulators (and an acquirer) that durable silencing isn't buying durable off-target risk?
6. Tune is ~12 months ahead in the identical indication and mechanism. Why does second place win here, and what is your realistic path to a differentiated pharma partnership if Tune signs the marquee deal first?
7. What is the HBV/HDV coinfection development plan and timeline, given HDV is a smaller but higher-unmet-need, potentially faster-to-differentiation population?
8. Of the ~$360M invested across Chroma and Nvelop pre-merger, what did that capital buy, and what specifically changes with the leaner combined structure?
9. What are the two or three most advanced extrahepatic / eVLP programs, their target indications, and their expected IND timing — i.e. is there a credible second clinical asset before 2028?
10. What is your freedom-to-operate position on the foundational epigenetic-editing IP relative to Tune, Epic Bio, and the broader CRISPRoff patent estate?
11. If GSK's bepirovirsen (re-dosable ASO, ~19% functional cure) reaches market first, how does a one-dose editor priced at gene-therapy economics compete on payer value?
12. What CDMO / manufacturing dependencies sit on the critical path to the readout, and what is the single-point-of-failure risk to the late-2026 timeline?
13. Why initiate first-in-human in Hong Kong, New Zealand, and the UK rather than the US/FDA — cost, speed, HBV epidemiology, or a regulatory-risk consideration — and when does an FDA IND follow?
14. What would a go/no-go decision on CRMA-1001 look like — what readout result would make you stop the program, and what would make you accelerate a registrational study?
15. With Marrazzo (Spark→Roche) as chairman, is the explicit strategy to build-to-acquisition, and what data package does a Roche-grade acquirer need to see to move?

---