Absci Corporation

2.1 ABS-101 (anti-TL1A, IBD) — from lead asset to outlicensing in six months

ABS-101 is a subcutaneous anti-TL1A monoclonal antibody engineered for extended half-life (see Section 4.1 for the Morgan Stanley counter-read that the same data "trailed next-generation competitors"). Absci built it to compete in inflammatory bowel disease on dosing convenience, not target novelty. Absci Pty Ltd sponsored the first-in-human Phase 1 at Nucleus Network in Melbourne (ACTRN12625000212459); the first participant dosed on 12 May 2025 C-ABS101-01. The US IND cleared but never opened a US site C-ABS101-04; the single Australian site narrowed US-investor visibility on the program.

The trial ran a single-ascending-dose design: four subcutaneous cohorts (150, 300, 600, 1000 mg) plus an IV bridge cohort at 300 mg, eight participants per cohort, target n=40 C-ABS101-02C-ABS101-03. The ANZCTR registry has not updated since 2025-07-22, when accrual stood at 15 of 40 V_NEW-009. The November 2025 interim readout therefore rested on fewer than twenty participants.

The interim disclosure described "extended half-life as compared to first-generation TL1A antibodies" with "no apparent impact of ADA on PK" H-043. The framing omitted the second-generation programs that now define the class — duvakitug (Sanofi/Teva, SC Q2W, $1.5B Sanofi license) and afimkibart (Roche, t1/2 18-19 days, $7.1B Telavant acquisition). Morgan Stanley read the same numbers and downgraded ABSI from Overweight to Equal-weight on 2026-01-08, cut its price target from $5.80 to $4.32, and removed ABS-101 from its model H-044H-046. The MS note states ABS-101 "showed half-life characteristics not in line with next-generation TL1A comparators" H-045.

Six days after the interim release, Absci announced the deprioritization: "the Company has made the strategic decision not to initiate additional later-stage development trials for ABS-101 internally at this time" V_NEW-009. Absci now seeks a partner. The TL1A class precedent on half-life-engineered programs is unhelpful: BB-TL1A-VIAL-HLE (NCT07029971, sponsor Vial Australia — not Boehringer Ingelheim) terminated at interim (n=16) on PK endpoint failure C-TL1A-V01. Animal-to-human PK translation in HLE-engineered TL1A is not guaranteed.

The class is also commercially crowded.

Merck paid $10.8B for Prometheus in April 2023 to secure tulisokibart B-016.

Roche paid $7.1B upfront plus $150M in milestones for Telavant's afimkibart in October 2023 E-011.

Sanofi licensed duvakitug from Teva for EUR 469M upfront plus EUR 940M milestones in October 2023 B-020.

AbbVie paid FutureGen $150M upfront plus $1.56B milestones for the preclinical FG-M701 in June 2024 F-008.

Boehringer Ingelheim entered the class on 2026-01-27 by licensing Simcere SIM0709 (TL1A x IL-23p19 bispecific) for EUR 1.058B V_NEW-006 — a re-entry, not an exit.

ABS-101 enters fifth or sixth in line with one differentiation story the analyst class has rejected H-045.

2.2 ABS-201 (anti-PRLR, AGA + endometriosis) — sole internal asset, second mover

ABS-201 is an AI-designed anti-prolactin receptor monoclonal antibody from Absci's Origin-1 platform. The company first disclosed the target and lead indication at its December 2024 R&D Day; prior quarters had described it only as "an undisclosed dermatology target." First-in-human dosing occurred on 3 December 2025 in the HEADLINE Phase 1/2a trial (NCT07317544), a quadruple-masked, 227-participant study at Sinclair Dermatology and Nucleus Network in Melbourne C-ABS201-01C-ABS201-02. The design runs four to six SAD IV cohorts (150-1800 mg, a twelve-fold dose span) followed by three to four MAD subcutaneous cohorts (300-1200 mg) C-ABS201-03.

Two design facts qualify the marketing language. First, the registered primary endpoint is safety alone; hair-growth measures (TAHC, TAHW, TAHD, IGA, self-assessment) sit in the otherOutcomes tier — neither secondary nor primary C-ABS201-05. Absci's "powered to demonstrate human proof-of-concept" framing for the MAD portion does not match the registry C-ABS201-05. Second, the pipeline page shows ABS-201 AGA "Phase 1/2a at 85%" and ABS-201 endometriosis "Phase 1 at 85%" although three of four-to-six SAD cohorts have completed and no endometriosis trial has been registered H-038; Phase 2 endometriosis initiation guidance is Q4 2026. Progress bars overstate program momentum.

The mechanism-priority claim fails the public record. Hope Medicine's HMI-115 — licensed from Bayer in 2019 — published Phase 2 endometriosis results in Lancet Obstetrics, Gynaecology, and Women's Health on 2025-11-05 (NCT05101317, n=108 completed, randomized 1:1:1:1 to 60/120/240 mg HMI-115 vs placebo, SC Q2W x 12 weeks) and registered Phase 3 endometriosis on 2025-12-29 (NCT07318688, n=540) V_NEW-005C-PRLR-H01. HMI-115's Phase 2 male AGA trial (NCT06118866, n=192, single-site Peking University People's Hospital) primary-completed on 2024-08-26 and remains unpublished as of report date V_NEW-005. ABS-201 is second mover in both indications, with HMI-115 sitting one phase ahead in endometriosis and one published readout ahead in mechanism validation.

The preclinical package behind ABS-201 rests on company disclosures, not peer-reviewed publications. Non-human primate studies showed greater than 90% subcutaneous bioavailability and projected Q8W-Q12W dosing in humans C-AGA-P01. Absci targets a high-concentration formulation at 150 mg/mL or above C-AGA-P01. In December 2025 the company described "statistically superior hair regrowth compared to minoxidil in a preclinical mouse model"; no effect size and no peer-reviewed publication have followed, and the telogen-to-anagen-transition mouse AGA model translates poorly to human androgen-driven follicular miniaturization H-018. March 2026 added human ex vivo data showing STAT5 phosphorylation inhibition and anagen prolongation in isolated follicles, mechanistically coherent with the published PRLR follicle literature G-008.

2.3 ABS-301 (immuno-oncology, undisclosed target)

ABS-301 remains at Lead ID stage on the pipeline page H-038. Absci has not disclosed the target since its December 2024 unveiling; a Q3 2025 update reported completion of "the first in vivo target validation study" with signaling-pathway-driven anti-tumor response H-043. December 2024 guidance called for candidate selection in 1H 2025; sixteen months later the asset still shows Lead ID H-038. The timeline slipped without acknowledgement.

2.4 ABS-501 (anti-HER2)

ABS-501 is an AI-designed anti-HER2 antibody from the same platform, displayed as Candidate at 0% H-038. Absci describes it as having "novel epitope interactions, affinity equal or greater than trastuzumab in preclinical, efficacious against trastuzumab-resistant xenograft." The international application WO2025122885A1 was filed 8 December 2023 with no granted US composition-of-matter patent as of report date B-011. Trastuzumab is a twenty-five-year-old benchmark; preclinical affinity parity is the table stakes for entry, not differentiation.

2.5 History — from SoluPro to Origin-1

Absci began in a 400-square-foot basement lab in 2011 D-001. Founder Sean McClain holds inventorship on the original SoluPro/SoluPure cytoplasmic expression patent (US12157766B2, priority 2012-08-05, granted 2024-12-03) B-003, and the earliest Absci-assigned US patent (US9617335B2) granted in 2017 B-008. The original IP base was microbial protein expression. Absci pivoted to generative-AI antibody design later, accelerated by the June 2021 Totient acquisition (US12285484, US12134655) B-002B-004. The Origin-1 platform brand launched in 2026; its predecessor brand was "Integrated Drug Creation."

3.1 TL1A (target validation score: 85)

TL1A drives T-cell activation and fibroblast-mediated tissue damage in inflammatory bowel disease E-006. Three independent lines of evidence anchor the target: GWAS signals in TNFSF15 associated with both ulcerative colitis and Crohn's disease; elevated TL1A expression in inflamed gut mucosa; preclinical knockout and antibody-blockade models that attenuate colitis E-006. Clinical validation has now replaced preclinical inference C-TL1A-T02. Tulisokibart's Phase 2 ARTEMIS-UC posted 26% clinical remission versus 1% placebo at 12 weeks (NEJM 2024) C-TL1A-T02E-005E-006; duvakitug RELIEVE reported 48% UC remission at 900 mg vs 20% placebo C-TL1A-D03; Roche's afimkibart TUSCANY-2 reached 36% remission at Week 56 E-011.

The class's commercial validation matches its biological validation E-005E-006.

Merck paid $10.8B for Prometheus to secure tulisokibart B-016.

Roche paid $7.1B for Telavant six months later E-011.

TL1A scores 85 — biologically established E-006, clinically proven C-TL1A-T02, commercially priced B-016F-008.

ABS-101 inherits all of this validation. Its question is not whether TL1A is a real target but whether a fourth-place HLE-engineered program can earn Phase 2/3 capital from a licensee whose closest comparator (BB-TL1A-VIAL-HLE) just terminated at n=16 C-TL1A-V01.

3.2 PRLR (target validation score: 42)

Prolactin receptor signals through JAK2-STAT5 in hair follicles, breast, ovary, bone, and pituitary G-008. Prolactin engagement at PRLR drives the catagen phase of the hair cycle; PRLR blockade prolongs anagen in mouse models G-008. Absci's human ex vivo data confirm STAT5 inhibition and anagen prolongation in isolated follicles. In endometriosis, PRLR signaling is upregulated in ectopic lesions and drives inflammation independently of the GnRH axis that current therapies target.

Clinical validation rests on one program: HMI-115 V_NEW-005.

The Lancet OGWH paper (2025-11-05; n=108 completed) reported a -41.57% LSM change from baseline in dysmenorrhea NRS at week 12 in the HMI-115 240 mg arm, placebo-adjusted -22.96%, p=0.036 V_NEW-005.

Only the 240 mg dose reached statistical significance against placebo; the 60 mg arm gave p=0.409 and the 120 mg arm gave p=0.131 V_NEW-005.

Non-menstrual pelvic pain reduction reached roughly 52% LSM versus placebo V_NEW-005.

The trial reported no sex-hormone disruption and no change in BMD, estradiol, LH, FSH, or progesterone V_NEW-005.

An AGA Phase 1b (Hope Medicine, n=12 males, open-label, single-arm) reported a +14 hairs/cm-squared mean non-vellus TAHC gain over 24 weeks per the January 2024 release V_NEW-005; this is not a Phase 2 benchmark.

Bayer's own mouse endometriosis data (Otto et al., 2022) showed efficacy equivalent to faslodex and cetrorelix without anti-estrogenic effects G-008.

Two cautions temper the validation. Novartis's anti-PRLR LFA102 failed Phase 2 in breast cancer — a different indication, but the only prior anti-PRLR in a large efficacy study V_NEW-005. Bayer's own BAY 1158061 completed Phase 1 in 2018 (PMID 29806538) with favorable PK and posted no Phase 2 registry entry in the eight years before HMI-115 returned the molecule to the clinic V_NEW-005. The silence is unusual for a drug with positive preclinical efficacy data. PRLR scores 42 — the mechanism is biologically supported, one competitor has generated supportive Phase 2 endometriosis data at one of three doses V_NEW-005, and no anti-PRLR therapy is approved.

3.3 Origin-1 platform claims

Absci's platform asserts de novo antibody design against "zero-prior" epitopes — targets without published protein-protein complex structures or homologs above 60% sequence identity G-005. The 2026 bioRxiv preprint (Levine et al.) reports developable binders for four of ten targets (COL6A3, AZGP1, CHI3L2, IL36RA) with cryo-EM atomic accuracy (3.0-3.1 Angstrom, DockQ 0.73-0.83) and an IL36RA binder optimized to 104 nM G-005. Absci characterizes this as "the first demonstration of de novo design of full-length, monoclonal antibodies against 'zero-prior' epitopes with atomically accurate complex structures and functional activity" G-005.

Three constraints bound the platform claim.

The 4/10 success rate is a 40% hit rate; six of ten targets failed to produce developable binders G-005.

All 54 authors on the Origin-1 preprint are Absci employees or contractors; no independent group has published a validation G-005.

The Baker Lab's RFdiffusion (Bennett et al., Nature 2025-11-05) demonstrated atomically accurate de novo design of antibodies against zero-prior epitopes — TcdB, influenza HA, RSV, SARS-CoV-2 RBD, IL-7Rα — at 0-2% per-design success rates G-025.

The peer-reviewed Baker Lab demonstration narrows Absci's "first demonstration" preprint claim, and Xaira Therapeutics — co-founded by Baker, capitalized at $1B+ E-028F-074 — is the venture-backed catch-up vehicle.

No peer-reviewed PubMed paper describes the de novo antibody-design platform method itself — five PubMed Absci-affiliation hits exist, but each covers histology AI (Iyer et al., Nat Med 2024, primary affiliation PathAI), HIV antibody biology (Spencer et al., Nat Comm 2022, primary OHSU), an EDI workshop report (Plant Direct 2022), QuickIsoSeq RNA-seq (Methods Mol Biol 2021, primary Pfizer), or pre-pivot Vibrio biology (Appl Environ Microbiol 2020) V_NEW-009.

None describes the Origin-1 method [cid:V_NEW-009; verifier V3-headline-9].

The 2023 flagship anti-HER2 de novo bioRxiv preprint (Shanehsazzadeh et al., DOI 10.1101/2023.01.08.523187) remains unreviewed three years after posting G-005.

Absci's most distinctive claim — zero-prior de novo design — has not been tested against a human clinical target G-005.

ABS-201 uses PRLR, which has extensive prior structural knowledge G-008.

4.1 ABS-101 Phase 1 — the half-life comparison Absci did not make

The November 2025 disclosure described "extended half-life as compared to first-generation anti-TL1A competitor programs, with no apparent impact of ADA on PK" H-043. Three things the framing omits:

1. The numeric half-life of ABS-101. No company disclosure provides t1/2 in days. Morgan Stanley filled the gap on 2026-01-08: "trailed next-generation competitors" H-045.
2. The comparator. "First-generation anti-TL1A" omits duvakitug (SC Q2W) C-TL1A-D03 and afimkibart (t1/2 18-19 days) E-011 — the second-generation programs that define the class.
3. The sample size. ANZCTR accrual stood at 15 of 40 on 2025-07-22 with no subsequent update V_NEW-009; the interim release rested on fewer than twenty participants.

A December 2024 R&D Day disclosure reported "reduced internalization of TL1A complexes in in vitro THP-1 immunogenicity tests compared to a competitor molecule with a high clinical anti-drug antibody (ADA) rate" H-043. The unnamed comparator is likely PF-06480605 (the tulisokibart precursor), flagged in the literature as a high-ADA molecule (no primary source — author estimate; specific ADA rate not verified in canonical). That context validates the engineering choice but does not translate to a head-to-head ADA comparison in humans. The 13-week GLP toxicology reported "no treatment-related adverse findings during in-life phase and necropsy were observed; histopathology is pending" H-043. Absci has not updated the "histopathology is pending" qualification since December 2024.

The Phase 1 outlicensing decision followed the interim readout by six days V_NEW-009. Had the data cleanly beaten the second-generation competitors on head-to-head-comparable PK metrics, the narrative and the timing would likely have differed.

4.2 ABS-201 HEADLINE — three SAD cohorts and a proof-of-concept that is still to come

HEADLINE's three completed SAD IV cohorts have reported favorable emerging safety data and no serious adverse events C-ABS201-01C-ABS201-02. The proof-of-concept readout requires the MAD SC portion, which has not yet started. Company guidance calls for preliminary safety/tolerability/PK data in 1H 2026, interim POC in H2 2026, and full POC in early 2027 C-ABS201-03. The cadence is reasonable for a Phase 1/2a that began dosing on 3 December 2025 C-ABS201-01.

Two design features warrant note. The SAD IV dose range of 150-1800 mg spans twelve-fold; for a molecule destined for subcutaneous administration, the 1800 mg top IV dose reads as a supratherapeutic-exposure safety probe more than a clinical development step. The registered statistical plan does not power any hair-growth endpoint as primary or secondary — TAHC, TAHW, IGA, and self-assessment all sit in otherOutcomes C-ABS201-05. Any H2 2026 hair-growth result will be a marketing framing, not a registered endpoint hit.

4.3 Anti-anchoring on preclinical and TAM claims

Several Absci disclosures require calibration against the public record H-017H-018H-043:

• "80 million Americans with AGA." The figure includes all degrees of hair loss; the addressable moderate-to-severe subset is materially smaller. The Grand View Research global AGA TAM is $2.7B in 2023 (8.45% CAGR to ~$3.6B by 2030) V_NEW-011F-020 — the public benchmark, not Absci's $25B management figure F-024.
• "Approximately 9 million women in the U.S. with endometriosis" and "prevalent in up to 10% of women worldwide" H-017. Absci truncated the 5-10% range in its own December 2025 disclosure to the upper bound in subsequent messaging.
• "No FDA-approved disease-modifying therapy" for endometriosis. Orilissa (elagolix, 2018) and Myfembree (relugolix combination, 2022) carry FDA approval for endometriosis-associated pain H-017. Absci's blanket framing contradicts the FDA label record; the defensible reading is "no non-hormonal disease-modifying therapy."
• "Statistically superior hair regrowth compared to minoxidil in a preclinical mouse model." No effect size, no peer-reviewed publication, and a model with weak human translation G-008.
• "2-3x extended half-life as compared to antibodies in clinical development" (August 2024 ABS-101 NHP data) H-043. Absci did not name the comparator antibodies. No peer-reviewed corroboration.
• Pipeline-page progress bars. ABS-101 "Phase 1 at 85%" despite the November 2025 internal halt H-038; ABS-201 AGA "Phase 1/2a at 85%" with three of four-to-six SAD cohorts dosed; ABS-201 endometriosis "Phase 1 at 85%" with no clinical trial registered. The website overstates asset momentum.

4.4 The peer-review gap

Absci has published no peer-reviewed journal article describing its de novo antibody-design platform method G-005. The five PubMed Absci-affiliation hits cover unrelated topics — three list Absci as a secondary affiliation, two predate the AI pivot. The 2023 Shanehsazzadeh preprint and the 2026 Origin-1 preprint both remain unreviewed [cid:V_NEW-009 verifier V3-headline-9]. For a company raising capital and signing partnerships on platform superiority, the continued reliance on bioRxiv-only evidence constrains the science narrative.

5.1 TL1A class safety signal score: 22 (LOW)

The TL1A class has been de-risked clinically. Tulisokibart's Phase 2 ARTEMIS-UC reported a clean safety profile comparable to placebo C-TL1A-T02. Duvakitug and afimkibart Phase 2 data showed no class-defining safety signals C-TL1A-D03E-011. FAERS pharmacovigilance reporting on the class has not generated disproportionality signals against IBD-biologic reference drugs. ABS-101's interim "no serious adverse events" across roughly fifteen participants H-043 is consistent with the class. The score reflects a mechanism understood in inflammation, three Phase 2/3 programs with clean readouts, and no regulatory holds E-005E-006.

5.2 PRLR class safety signal score: 58 (MODERATE-HIGH)

The PRLR safety picture is less mature. The knockout phenotype in mice is documented: female homozygous mice are infertile and lack mammary development; hemizygous females cannot lactate after their first pregnancy; pituitary prolactin levels rise 30-100-fold with pituitary hyperplasia in some cases; bone formation falls in both sexes at all ages G-008. Each phenotype predicts a potential on-target liability for chronic anti-PRLR administration in humans.

The HEADLINE Phase 1/2a protocol (NCT07317544) does not specify DEXA bone-density measurement or bone-turnover biomarker monitoring C-ABS201-04. In a chronic-use asset aimed at AGA — a non-life-threatening indication competing with cash-pay generics — the absence of bone-turnover monitoring in the Phase 1/2a readout is a protocol gap Phase 2 design must close. HMI-115's Lancet Phase 2 reported no sex-hormone disruption and no fertility signal V_NEW-005; Bayer's BAY 1158061 Phase 1 posted favorable PK without disclosed safety concerns V_NEW-005. Neither program has generated long-term (>12 months) human safety data. The score reflects on-target phenotype predictions humans have not yet ruled out, plus the absence of bone-turnover monitoring in the registered protocol G-008C-ABS201-04.

5.3 Implications for development

For ABS-101, the low class safety signal supports outlicensing: a partner faces no unusual regulatory headwinds H-043. For ABS-201, the moderate-high class signal means Absci cannot rely on safety alone to differentiate G-008. Efficacy must separate from HMI-115 V_NEW-005 and from minoxidil/finasteride to justify chronic use in a cosmetic-adjacent indication. Endometriosis changes the calculus: pain severity is real, the unmet need is real, and the safety bar is higher in theory but lower in practice.

6.1 Nine granted US patents verified (one short of the 10-K)

Independent Google Patents and USPTO verification confirms nine granted US patents assigned to Absci Corp or Absci LLC as of report date B-015:

The 10-K's "ten issued U.S. patents" B-015 exceeds the nine verifiable via public registries by one.

The gap may reflect a recently granted tenth indexed after report cutoff, or a 10-K counting inconsistency tied to subsidiary entities (Totient UK Ltd., Absci Pty Ltd., Absci GmbH) B-015.

6.2 ABS-201 has no granted US composition-of-matter patent

The most material IP gap sits on the only internally-advancing asset. ABS-201 carries no publicly discoverable granted US composition-of-matter patent as of report date B-021, although HEADLINE Phase 1/2a began dosing on 3 December 2025 C-ABS201-01. The international application is pending, but the public record does not yet confirm a US grant. An AI-designed antibody entering clinical development without a granted US CoM patent in its destination indications carries freedom-to-operate risk that sharpens against the Bayer prior art described below.

6.3 The ABS-101 freedom-to-operate concern (Merck/Prometheus)

US11999789B2 (Cedars-Sinai, Prometheus/Merck) — "Humanized antibodies to TL1A" — granted 4 June 2024, expires 23 October 2040 B-016.

The continuation US20240327532A1 (filed 2024-04-24, published 2024-10-03) may broaden into genus or functional claims; monitor for grant B-017.

Absci's ABS-101 CoM patent US12473370B2 covers six specific CDR sequences B-001; no independent overlap analysis with the Merck/Prometheus claims has been published.

The outlicensing pivot reduces Absci's direct FTO exposure but does not eliminate it as a constraint on the deal value Absci can extract V_NEW-009.

Paragon Therapeutics added a fifth granted US anti-TL1A patent (US12466890B1, granted 2025-11-11) B-018, one week before Absci's grant — the class IP landscape is crowded.

6.4 The ABS-201 freedom-to-operate concern (Bayer / Hope Medicine)

WO2019011719A1 — "Prolactin receptor antibody for male and female pattern hair loss," Bayer IP GmbH, priority 10 July 2017 — is the primary FTO concern for ABS-201 B-021. Bayer licensed the asset to Hope Medicine, which developed HMI-115 and completed Phase 2 AGA in August 2024 (n=192) V_NEW-005. Bayer's earlier WO2011069799A1 expands the prior-art surface B-021. Absci's ABS-201 may design around these claims on specific sequence and epitope grounds, but a competitor with licensed Bayer IP is already running Phase 2 trials in ABS-201's two target indications V_NEW-005. The competitive landscape and the FTO landscape are the same landscape.

6.5 Pending method claims — the core AI platform IP is not yet granted

US20230268026A1, "Designing biomolecule sequence variants with pre-specified attributes," filed 7 January 2022, represents the core generative-AI method claim and remains pending B-014. Absci holds no granted US method patent for its de novo antibody design algorithm. WO2025122885A1 (ABS-501 anti-HER2) B-011, WO2024040020A1 (ACE Assay) B-012, and WO2025144700A1 (nanobody screening) B-013 represent the global filing strategy but do not become enforceable until national-phase grant.

The Generate Biomedicines US12020776B2 patent — earlier framed as a "direct overlap" risk — addresses a different problem at the claim-1 level. Generate's claim 1 covers a humanization workflow using MHC-II residue-wise immunogenicity reweighting V_NEW-013. Absci's pending US20230268026A1 claim 1 covers a binding-characteristic prediction model trained on measured binding data B-014. The two claim-1 texts do not read on each other element-by-element. The current assignee on Generate's grant is Flagship Pioneering Innovations VI Inc, not Generate Biomedicines per se V_NEW-013. A complete Origin-1 FTO opinion against Generate's 71-application portfolio would require dependent-claim and specification-level analysis by a registered patent attorney; this DD does not constitute that analysis.

6.6 Portfolio expiration window

The 10-K reports an expected expiration range of August 2033 through December 2044 B-015. The earliest-expiring patents — the 2012-priority SoluPro family — protect platform manufacturing more than antibody composition B-008B-009. The CoM patents protecting specific assets (US12473370B2 for ABS-101 to 2044-10-02; pending for ABS-201) extend to 2044 B-001B-021. The gap between platform-IP and asset-IP expiration narrows long-horizon defensibility against biosimilar entry.

7.1 Executive team

Sean McClain — Founder and CEO has run Absci since inception in August 2011 D-001D-002. He holds a BS in Molecular and Cellular Biology from the University of Arizona D-002. No prior exits; no CEO history outside Absci. His 2024 total compensation reached $9,786,988, driven by a $5,491,250 Monte Carlo PSU grant — 3.2x his 2023 total D-003. Across fourteen years he has taken one company from basement lab through IPO into clinical stage. The pattern of long tenure without governance conflicts is favorable; the absence of prior exits leaves his operating-CEO track record unbenchmarked outside Absci.

Zach Jonasson, PhD — CFO and Chief Business Officer assumed the role on 31 August 2023 after serving as an independent director from October 2020 to December 2023 and on the Absci LLC predecessor board from April 2016 D-004D-005. He co-founded Phoenix Venture Partners and Convergent Ventures. His 2024 total compensation reached $2,832,851 D-006. Two governance facts require disclosure. Phoenix Venture Partners II LP (Jonasson's prior firm) is a Schedule 13G 10%+ beneficial owner of ABSI (9,354,695 shares filed 19 February 2025) and was disclosed as a delinquent Section 16 filer in the 2025 proxy along with Jonasson himself — Phoenix's late filing covered 3,100,000 shares V_NEW-003. Jonasson moved from independent director into an executive officer at a company where his former fund is a 10%+ holder.

Ransi Somaratne, MD, FACC, MBA — Chief Medical Officer started 3 March 2026 D-007D-008. He joined from Vertex Pharmaceuticals, where he served as SVP Clinical Development (Jan 2023 - Mar 2026); earlier roles at BioMarin (Roctavian hemophilia A gene therapy), Amgen (Repatha PCSK9 Global Development Leader), and short CMO tenures at Epirium (1 yr 8 mos), Spring Discovery (1 yr), and Endpoint Health (7 mos) D-009. MD Albany Medical College; MBA UNC Kenan-Flagler; Cardiology Fellowship Kaiser Permanente LA; BS Mechanical Engineering UC Santa Barbara. Somaratne is Absci's first full-time CMO, hired ten months after ABS-101 entered Phase 1 and three months after ABS-201 first dose D-007. His inducement grant — 650,000 options at $2.57 (the 3 March 2026 close), 4-year vest — sits 14% below the July 2025 offering price of $3.00. LinkedIn shows a concurrent role at Iles & Somaratne Biopharmaceutical Consulting LLC (Sep 2021 - present) the appointment 8-K did not disclose D-007.

Andreas Busch, PhD — Chief Innovation Officer retires effective 31 March 2026 and transitions to SAB co-chair on 3 April 2026 with a 2-year scientific advisor agreement at $25,500/year cash plus 22,800 options + 5,800 RSUs and an exercise-window extension on 2,489,290 vested options D-012. The 8-K states the resignation "was not the result of any disagreement with the operations, policies or practices of the Company" D-012. Busch's $381,000 open-market purchases over the trailing twelve months — including $229,000 on 2026-03-12 during the retirement transition — signal personal conviction from a departing officer D-012V_NEW-001. His 2024 total compensation reached $2,644,420, paid via Swiss subsidiary Absci GmbH at a 1.14 USD/CHF average rate D-013.

Amir Shanehsazzadeh — Chief AI Officer ascended to the CAIO title in April 2025, replacing Amaro Taylor-Weiner. Harvard AB Mathematics and SM Statistics. Shanehsazzadeh is an inventor on the ABS-101 CoM patent US12473370B2 B-001 and first author of the 2023 flagship de novo antibody bioRxiv preprint G-005. Absci did not disclose the CAIO promotion via dedicated 8-K, the title is administrative not Section 16, and no Form 4 filings exist under his name — disclosure obscurity at an AI-branded company. His tenure began as an AI Scientist in June 2022 — a fast internal track for the central thesis.

Shelby Walker, JD — Chief Legal Officer joined June 2024 from Korro Bio GC/Corporate Secretary (May 2023 - June 2024); prior SVP Intellectual Property at CRISPR Therapeutics (March 2018 - April 2023) and GC at Ginkgo Bioworks (May 2016 - March 2018) D-014. JD/LLM UNH Franklin Pierce School of Law; dual MS Bioscience Regulatory Affairs and Biotechnology Johns Hopkins; BS Biotechnology WPI. Deep IP track record for a company whose valuation depends on IP.

Todd Bedrick, CPA — Chief Accounting Officer has served since June 2022 D-010. He acquired 124,200 shares and disposed of 36,173 net over the trailing twelve months V_NEW-001 — small but directionally consistent with insider bullishness. Bedrick signs the 10-K as Principal Accounting Officer (a Section 16 capacity) but the DEF14A 2025 executive officer biography roster excludes him — inconsistent disclosure across filings.

7.2 Board of Directors (seven members)

The board fixed at seven seats on 7 July 2025 when Mary Szela joined Class III D-023. Current composition: McClain (CEO/Class III), Pangalos (Class II, since Jan 2024) D-021, Rabinovitsj (Class II, since Nov 2022) D-022, McGinnis (Class I, audit chair, since Aug 2020) D-020, van Houten (Class I, board chair, since June 2023) D-019, Sirosh (Class III, since Jan 2022), Szela (Class III, since July 2025) D-023D-024. Class split is 2/2/3, not 3/3/1 as some earlier ledgers reported.

Sir Mene Pangalos, PhD served as EVP BioPharmaceuticals R&D at AstraZeneca from January 2019 to March 2024 D-021. He joined the Absci board in January 2024, one month after the December 2023 AstraZeneca $247M collaboration announcement V_NEW-007; the public record discloses no firewall arrangement between his AZ R&D role and his Absci director candidacy. His trailing twelve-month open-market purchases of $460,548 (170,880 shares) are the largest among insiders V_NEW-001. Biogen board service runs in parallel.

Mary Szela joined 7 July 2025 with a 178,400-option inducement grant D-023D-024. CEO of TriSalus Life Sciences since January 2018; previously CEO of Novelion Therapeutics and Melinta Therapeutics; earlier ran Abbott Laboratories' $8B US pharma business D-024. Also on the boards of Kura Oncology, Prometheus Biosciences (the company Merck acquired for $10.8B), Omega Therapeutics, and Senda Biosciences. Szela's LinkedIn shows "Board Member, Absci, Jul 2025 - Feb 2026" with an end date, but she signed the 10-K on 2026-03-24 as Director and TriSalus DEF14A confirms current Absci service; the LinkedIn end date is a self-update error, not a departure D-016D-023.

Frans van Houten, MSc — former CEO of Royal Philips (March 2011 - October 2022) D-019. Novartis AG independent director since February 2017. The 10-K formalizes his title as "Chair of the Board" D-016D-019. Audit committee financial expert. His trailing twelve-month open-market purchases of $148,800 add to the insider net V_NEW-001.

Karen McGinnis, CPA — ex-VP and Chief Accounting Officer at Illumina (November 2017 - April 2021) D-020. Audit committee chair; designated audit committee financial expert D-016. iRhythm Technologies board joined July 2025.

Daniel Rabinovitsj — VP at Meta since August 2018 D-022; prior COO Ruckus Wireless, SVP Qualcomm Atheros D-016. Tech compensation norms differ from pharma peer-group benchmarks — relevant for a director who chairs the compensation committee.

Joseph Sirosh, PhD — CEO of CreatorsAGI Inc; former VP Amazon Alexa Shopping (October 2022 - November 2023); prior CTO Compass, CVP and CTO of AI at Microsoft (July 2013 - December 2018) D-016. AI-sector board bridge.

Amrit Nagpal (Managing Director, Redmile Group LLC) resigned from the board on 25 April 2025 with no disclosed disagreement D-016. Redmile is a healthcare-focused hedge fund; the departure of a Redmile MD reduces dedicated-biotech-fund engagement.

7.3 Insider open-market activity — net +$1.017M with zero discretionary sales

Six trailing-twelve-month open-market purchases by four insiders total $1,017,348.40 across 370,880 shares V_NEW-001: Pangalos $460,548, Busch $381,000, van Houten $148,800, Bedrick $27,000 V_NEW-001. Two prior-flagged "sales" — McClain 26,761 shares and Jonasson 17,496 shares, both 2026-02-02, code S — are RSU tax withholdings per the explicit Form 4 Note 1: "Amount reported represents the number of shares withheld by the Issuer to cover the tax withholding obligation in connection with the vesting of these restricted stock units and does not represent a discretionary trade by the reporting person" V_NEW-002. Net 12-month insider activity: $1.017M of buying, zero discretionary selling. The signal is bullish, led by directors and the retiring CIO, none of whom needed to commit personal capital and all of whom did.

7.4 Scientific Advisory Board and endometriosis advisory board

The SAB seats Andreas Busch (co-chair post-retirement) D-012, Mene Pangalos, Iain McInnes (Vice Principal, University of Glasgow), Hubert Truebel (CMO AiCuris), and Victor Greiff (Associate Professor, University of Oslo). The March 2026 endometriosis advisory board added Hugh Taylor (Yale), Linda Giudice (UCSF), Gaurang Daftary (Inception; Yale/Mayo), Steven Young (Duke), Zaraq Khan (Mayo), and Joan-Carles Arce (Repronovo) — credible academic coverage for an indication Absci has not yet registered a trial in.

7.5 Team summary

The bench has the credentials it needs: a commercially seasoned CMO (Somaratne), a finance/BD operator in the CFO role (Jonasson, with the Phoenix relationship flag), an IP heavyweight as CLO (Walker), and an in-house AI lead (Shanehsazzadeh). The board mixes pharma commercial experience (Szela), Big Pharma R&D experience (Pangalos), tech/AI experience (Sirosh, Rabinovitsj), and two audit-committee financial experts (McGinnis, van Houten). Two simultaneous C-suite transitions — a new CMO and the CIO's retirement — during a clinical-ramp period D-007D-012 invite execution-risk scrutiny. The net-bullish insider purchase pattern, led by the directors and the outgoing CIO, signals personal conviction that outlasts the reorganization.

14.1 rNPV and scenario table (ABS-101 + ABS-201 combined, unrisked peak-sales)

Important caveat. The probabilities and IRRs in this table are this report's scenario construction, not primary-source facts. They rest on published industry base rates plus qualitative adjustments. Each number is a judgment call. (rNPV scenarios are author construct; success-rate base rates from BIO/BioMedTracker/Informa Clinical Development Success Rates 2011-2020 — no primary citation in canonical, analyst judgment.)

Evidence anchors for PoS inputs: Bull 0.25 = IBD-weighted antibody Ph1-to-approval upper quintile (BIO/BioMedTracker/Informa, Clinical Development Success Rates and Contributing Factors 2011-2020). Base 0.14 = industry-average antibody Ph1->approval (Informa 2011-2020). Bear 0.06 reflects late-entry penalty for ABS-101 (fifth+ TL1A entrant) and second-mover commercial-share penalty for ABS-201 PRLR behind HMI-115 V_NEW-005. Peak-sales ranges follow market-stream builds: ABS-101 $300M-$2.0B unrisked F-004; ABS-201 endometriosis $300-900M F-040; ABS-201 AGA $50-200M F-020.

Current market cap ~$400-450M (live snapshots 2026-04-10/27) E-001E-002 against implied base-case EV $300-450M: market is pricing close to base case. Bull case implies 2.0-3.0x upside; Bear case implies 65-80% downside. The probability shift from prior reports (Bear 0.30 from 0.25; Bull 0.15 from 0.20) reflects the updated short interest V_NEW-004, the registered HMI-115 Phase 3 endometriosis lead V_NEW-005, and the pure-cash mid-2027 exhaustion math V_NEW-014.

14.2 Cap table and dilution path

Cumulative capital raised: ~$216M pre-IPO A-059A-060A-061A-062A-063 + ~$225M IPO + $170M post-IPO A-025A-029A-032 = ~$611M.

Accumulated deficit $624.8M A-008 roughly equals total capital raised.

Implied: every dollar raised has been spent, with no retained earnings to cushion the next cycle A-008.

Implied pre-money today: ~$400-450M market cap E-001 minus ~$144M cash A-007 = enterprise value ~$256-306M E-004, in line with platform peers on EV/cash.

Cumulative dilution from IPO float (~90M shares) to current 155.4M shares A-011 exceeds 70%.

Burn rate $115M/year against $144M cash A-005A-007 matches 15-18 months hard runway absent further financing; the 1H 2028 guidance assumes another $50-90M of ATM draws between now and mid-2027 V_NEW-014.

14.3 Comparable transactions (last 30 months, same mechanism class or adjacent)

The TL1A deal cascade establishes a $1.5-10.8B band by stage F-008B-016V_NEW-006. ABS-101 at Phase 1 with a contested half-life narrative falls closest to the FutureGen floor H-045.

14.4 Forty-eight hour action plan

Decision now: MONITOR. The base-case EV matches current market cap E-001E-004; the Bull case requires two sequential wins (ABS-101 outlicense closes 2026 V_NEW-009 and ABS-201 AGA interim reads positive H2 2026 C-ABS201-03); the Bear case requires only one miss. Risk-adjusted return at current price is 6-12% IRR over three years — below biotech venture-equivalent hurdle rates. Wait for inflection data.

Triggers to reopen:

1. ABS-201 HEADLINE interim hair-growth readout, H2 2026 (TAHC delta vs baseline, with or without placebo separation) C-ABS201-03.
2. HMI-115 Phase 3 endometriosis primary completion (NCT07318688 estimated 2028-03-01) or HMI-115 Phase 2 AGA topline (NCT06118866, completed 2024-08-26, still unpublished — 16 months past primary completion) V_NEW-005.
3. ABS-101 outlicensing announcement with disclosed upfront terms (confirms base-case floor) V_NEW-009.
4. Large-pharma platform deal for Origin-1 beyond target-specific collaboration V_NEW-007.
5. Capital raise priced below $2.00/share (signals distress or capitulation) V_NEW-008.

Technical DD gaps to close before any position:

1. ABS-101 Phase 1 numeric PK in days, with the next-generation comparator Morgan Stanley referenced — verify the "extended" framing or confirm the MS reading H-045.
2. ABS-201 NHP PK data behind the 90% subcutaneous bioavailability and Q8-Q12 week dosing claims C-AGA-P01.
3. Bone-turnover monitoring plan for chronic-dosing indications — the PRLR-KO bone phenotype must be addressed before endometriosis Phase 2 G-008.
4. Internal ABS-201 vs HMI-115 head-to-head preclinical comparison, if any exists V_NEW-005.
5. Peer-review status on the Origin-1 bioRxiv preprint and the 2023 flagship HER2 preprint G-005.

Reference calls (3-5):

• Clinical dermatologist running comparator AGA trials — is NCT07317544's MAD protocol adequately powered for TAHC to differentiate from finasteride, given hair-growth endpoints sit in otherOutcomes? C-ABS201-05
• IBD gastroenterologist with TL1A investigator experience — how do pharma licensors weigh Phase 1 half-life differentiation against three Phase 3 competitors, post the MS analyst readout? H-045
• Ex-Hope Medicine executive or ex-Bayer BAY 1158061 program — what stopped the Bayer program between 2018 Phase 1 and 2019 HMI-115 licensing? Is there a silent liability behind the eight-year gap? V_NEW-005
• Public biotech analyst covering AbCellera / Recursion — what multiples apply to platform-plus-clinical-asset biotechs with declining partner revenue? E-017E-019
• Endometriosis KOL (Hugh Taylor at Yale or Linda Giudice at UCSF, both on Absci's endometriosis advisory board) — what level of pain-score separation would make ABS-201 a credible second mover after HMI-115's 240 mg signal? V_NEW-005

Commitment window: 120-180 days. The ABS-201 HEADLINE interim readout in H2 2026 is the dominant catalyst C-ABS201-03; no action before then unless triggers 2 (HMI-115 readout) V_NEW-005 or 5 (capital raise below $2.00) fire first V_NEW-008.

14.5 Tier-3 claims whose resolution would flip thesis

• Claim: ABS-101 "extended half-life" interim Phase 1 framing H-043. Why not verified: company press-release framing only, no PK curves, MS analyst counter-read H-045. Resolution effect: numeric t1/2 >30 days supports outlicensing toward $1.7B floor; <20 days collapses the differentiation story.
• Claim: "$247M AstraZeneca" and "$650M Almirall" headline totals V_NEW-007A-028. Why not verified: upfront and near-term cash not separately disclosed; AZ deal has zero ct.gov registrations 27 months in V_NEW-007. Resolution effect: upfront-plus-near-term cash >$50M per deal supports the platform-value case; pure-milestone terms at current realized pace collapse the platform revenue thesis.
• Claim: ABS-201 Q8-Q12 week dosing extrapolated from NHP PK C-AGA-P01. Resolution effect: human PK supporting Q8W+ materially differentiates from HMI-115's Q2W SC; Q2-Q4W eliminates differentiation.
• Claim: Origin-1 "first demonstration of de novo design against zero-prior epitopes with atomically accurate complex structures" G-005. Why not verified: bioRxiv preprint with 54 internal authors; no peer review. Resolution effect: peer-review acceptance at Nature/Science/Cell/Nat Biotech converts the platform claim from marketing to scientific consensus.
• Claim: "At least one or more partnerships, including with a Large Pharma company, in 2026" A-025. Why not verified: guidance language; carries forward from 2025 guidance that slipped. Resolution effect: a Large Pharma platform license (not program-specific) in 2026 validates Origin-1 as a licensable asset; absence by YE2026 confirms platform-revenue plateau.

Distribution: BigBio AI Research — Public. All sources are public filings, peer-reviewed publications, and company disclosures. No pitch-deck or VDR content. Every digit traces to canonical.yaml.