Act I · Physical & Clinical Story
Chemistry on a Single Base Inside Living Chromatin
Why it matters: Many of the most severe Class I CFTR mutations—including W1282X (UGG → UGA), R553X (CGA → UGA), R1162X (CGA → UGA), and the canonical splice variant 1717-1G>A—were caused by a historical G·C-to-A·T transition on the coding or non-coding DNA strand. Adenine Base Editing (ABE) reverses that exact atomic change: it converts the target adenine (A) into inosine (I), which the cell's DNA polymerase reads as guanine (G), durably repairing the endogenous gene without cutting both strands of the DNA helix.
Directly rearranging atomic bonds within the living genome without fracturing the double helix has long represented the ultimate objective of genetic medicine. Traditional molecular scissors sever both strands of genomic DNA, triggering chaotic cellular repair programs that frequently introduce unwanted insertions and deletions across non dividing airway tissues. Reporting in Science Translational Medicine during 2026, Lucas Umbach and an international team established that precision base editor enzymes can correct severe channel mutations without cutting chromosomes. By unwinding a localized bubble in genomic DNA, the engineered base editor exposes a single target letter. Tethered chemical enzymes then convert that isolated letter into an alternate base that cellular machinery reads as guanine. This non destructive conversion durably restores the native genetic instructions. Every subsequent messenger transcript produced by the repaired cell directs the synthesis of authentic full length chloride channels.
Many severe genetic alterations that disable the human chloride channel involve a single incorrect chemical letter in the underlying genetic blueprint. In these cases, an accidental historical transition turned an original instruction into an early stop sign or destroyed an essential splicing boundary. Standard gene editing tools attempt repair by cutting completely through both strands of the genetic helix. Because mature airway cells rarely divide, they struggle to mend severed strands correctly without introducing destructive insertions and deletions. Base editors circumvent this hazard entirely. Instead of cutting the helix, a specialized search enzyme opens a small ten letter window along one strand. A linked conversion enzyme rearranges the atoms of the faulty target letter, changing it into an alternate compound that cellular copiers interpret as guanine. The genetic code is repaired durably without a single double strand break.
Everyday Analogy: Consider an expert archivist restoring a priceless parchment manuscript where a single typo alters the legal meaning of an ancient decree. An aggressive repairer might slice out the entire damaged paragraph with a scalpel, leaving a gaping hole requiring an awkward patch. Instead, the master conservator applies a microscopic droplet of reagent directly onto the ink of the offending letter. The chemical reaction selectively alters the errant letter into the correct character without weakening the underlying parchment fibers. An adenine base editor operates inside living chromatin with that same degree of chemical restraint. It unzips a five-letter window of DNA and rearranges the atoms of a single target base. The chromosomal backbone remains completely intact, avoiding double-strand breaks. However, this archival restoration analogy stops being accurate because living chromatin undergoes active mismatch repair cascades that do not occur on static parchment.
How the Two Sliders Work in Plain Words: Before evaluating the kinetic equations governing nucleotide deamination, the interactive workbench lets you explore the spatial constraints of the base-editing complex. The first slider sets the Target Adenine Protospacer Position, determining where the mutant base sits within the single-stranded DNA bubble relative to the enzyme tether. Editing efficiency peaks when the target base aligns with the catalytic sweet spot between positions four and eight. The second slider tunes the Narrow-Window Deaminase Specificity, adjusting the spatial sharpness of the tethered catalytic domain. Examine the readouts. Sharpening the editing window boosts the On-Target A·T → G·C Editing Yield past fifty percent while driving the Bystander Adenine Deamination Rate down below two percent. Crucially, the Double-Strand Break Indel Rate remains negligible under one percent because the nickase spares the chromosomal backbone. These combined efficiencies drive the Restored Epithelial Cl⁻ Transport readout above forty-five percent of normal physiological conductance.
In 2017, Gaudelli, Komor, David Liu and colleagues invented Adenine Base Editors (ABEs) by evolving the bacterial tRNA adenosine deaminase TadA to operate on single-stranded DNA and fusing it to a Cas9 D10A nickase. When guided to a premature stop codon such as W1282X (5'-UGA-3' on mRNA, pairing with 3'-ACT-5' on the template DNA strand), ABE deaminates the template-strand adenine opposite the U, converting it to inosine (read as guanine)—which restores the wild-type tryptophan codon 5'-UGG-3' cleanly and durably.
Non-pooled national registry denominators frame the clinical motivation for transition base editing without summing cohorts across borders or equating modulator ineligibility with clinical efficacy of preclinical genomic editors: CFFPR 2024 reports 92.3% genotype-eligible, 26,277 of 29,314 (89.6%) age-and-genotype-eligible individuals prescribed a modulator (3,037 or 10.4% eligible unprescribed), and 8.0% modulator-ineligible; CCFR 2024 reports 7.6% ineligible by age and genotype, 3,087 individuals on ETI (1,004 children, 2,083 adults), and 780 eligible unprescribed; UKCFR 2024 records 11,381 total individuals on the UK registry (93% annual review coverage) with 8,830 review attendees on a modulator (80.3% Kaftrio; 12.4% never used a modulator); and ACFDR 2025 (31 Dec 2025, ex-transplant) reports 5.8% adult cohort ineligibility (2,326 known eligibility; 85.6% prescribed, 8.6% eligible unprescribed) and 11.5% (170 children) pediatric cohort ineligibility (1,473 children; 75.0% or 1,105 prescribed, 13.4% or 198 eligible unprescribed). Whereas deep-intronic 3849+10kb C->T retains canonical exons for steric ASO masking (`aso_splice_model.py`; SPL84 under NCT06429176 — ClinicalTrials.gov does not evaluate safety or efficacy; listing on the registry does not denote government endorsement.), canonical 1717-1G>A requires direct A·T-to-G·C repair as demonstrated by Umbach and colleagues in 2026.
“Directly repairs chromosomal DNA so every future CFTR mRNA transcribed from that edited cell is 100% wild-type, while avoiding double-strand break (DSB) indels or chromosomal translocations.”
— Cross-Pillar Research Thread · Grounded in Umbach et al. (2026), Science Translational Medicine 18(846):eadw8886 — Functional correction of the untreatable CFTR 1717-1G>A mutation through mRNA- and sgRNA-optimized base editing (PMID 42018671).
Key Parameters, Readouts & Primary Anchor
| Quantity / Anchor | Pinned Baseline | Physical / Clinical Role |
|---|---|---|
| Target Adenine Protospacer Position (x_target) | 6.0 pos (PAM=21) |
Primary perturbation axis (3.0 to 9.0 pos (PAM=21)) |
| Narrow-Window Deaminase Specificity (1 / σ_W) | 1.4window sharpness |
Coupled co-perturbation axis (0.6 to 2.0window sharpness) |
| On-Target A·T → G·C Editing Yield & Bystander Adenine Deamination Rate | 52.4% Clean WT Alleles · 1.8% (Narrow TadA-8e/9) |
Baseline biophysical operating point |
| Primary Literature Anchor | DOI 10.1126/scitranslmed.adw8886 | Umbach et al. (2026), Science Translational Medicine 18(846):eadw8886 — Functional correction of the untreatable CFTR 1717-1G>A mutation through mRNA- and sgRNA-optimized base editing (PMID 42018671). · Kernel src/scripts/ptc-abe-simulator.js |
Act II · From Intuition to Governing Law
Gaussian R-Loop Window Profiles and Bystander Discrimination
Inside the Cas9 R-loop bubble (numbered 1 to 20 from the 5'-end of the protospacer, with the PAM at positions 21–23), the accessibility of each displaced nucleotide x_i to the tethered TadA active site follows a localized spatial profile W(x_i) = exp(−(x_i − x_0)² / (2σ_W²)) centered near x_0 ≈ 5.5. If a target adenine sits at position 6 while another adenine sits at position 8 (as occurs in several CFTR nonsense contexts such as W1282X or R553X), a wide-window hyperactive deaminase risks editing both adenines—converting the stop codon into a missense variant (bystander editing). Narrowing σ_W through active-site steric mutations (V106W, N108Q) or shifting the PAM register maximizes P_clean = P_edit(x_target) · Π_{j ≠ target} (1 − P_edit(x_j)).
The spatial and kinetic probability of adenine base editing across the non-target single-stranded DNA bubble is formalized by P_edit(x_i) = η_R-loop · exp(−(x_i − x_0)² / 2σ_W²) · k_deam / (k_deam + k_off). Here, η_R-loop characterizes the thermodynamic efficiency of Cas9 R-loop formation, governed by guide RNA–DNA hybrid stability minus the energetic penalty of unwinding the genomic double helix. The Gaussian spatial factor exp(−(x_i − x_0)² / 2σ_W²) models the steric reach of the flexible peptide linker tethering the TadA deaminase domain to the Cas9 N-terminus, with mean position x_0 ≈ 5.5 and window variance σ_W². Within this exposed single-stranded bubble, the probability that target nucleotide x_i undergoes catalytic deamination depends on the kinetic competition between the intrinsic hydrolytic deamination rate k_deam and the microscopic R-loop dissociation rate k_off.
Once target adenosine is deaminated to inosine, mismatch resolution dictates the durable genomic outcome. The Cas9 D10A nickase domain cleaves the non-edited (guide-paired) DNA strand, generating a site-specific single-strand nick approximately three nucleotides upstream of the PAM. This directional nick recruits endogenous host mismatch repair factors (MutSα/MutLα) and exonuclease 1 (EXO1), degrading the unedited strand while utilizing the inosine-containing strand as the pristine template. DNA polymerase δ resynthesizes the gap by pairing cytosine opposite inosine, effectively fixing an I·C intermediate that converts to a durable G·C base pair upon subsequent DNA replication. In our computational simulation kernel (src/scripts/ptc-abe-simulator.js), this multi-step model demonstrates that narrow-window TadA variants (such as ABE8e-V106W) steepen σ_W, restricting editing to target position 6 while suppressing bystander deamination at positions 4 and 8 below 2% across the CFTR locus.
Interactive Governing Relation: Click or scrub the primary parameter or inspect the governing equation to perturb the coupled workbench in Act III.
Four-Stage Mechanistic Progression
- Stage 1 (PAM Recognition & R-Loop Bubble Formation): SpCas9-NG or PAM-relaxed Cas9 nickase (D10A) binds the target locus and unwinds a ~20-nt RNA–DNA hybrid, exposing positions 4–8 of the displaced single-stranded DNA strand.
- Stage 2 (TadA Zinc-Mediated Hydrolytic Deamination): The evolved deoxyadenosine deaminase (TadA-8e / TadA-9) flips the target adenine into its Zn²⁺ active site, catalyzing proton-coupled nucleophilic attack to convert adenosine into inosine (A → I).
- Stage 3 (Non-Edited Strand Nicking & Mismatch Resolution): The Cas9 D10A HNH domain nicks the target (guide-paired, non-deaminated) DNA strand, biasing eukaryotic mismatch repair (MMR) to use the inosine-containing strand as the template—fixing an I·C → G·C base pair.
- Stage 4 (Bystander Window Control): Narrow-window TadA variants (such as ABE8e-V106W or TadA9) shrink the Gaussian activity width σ_W so neighboring adenines at positions 3 or 9 remain untouched.
Act III · Interactive Visual Experiment & Plain-Language Lens
Try It Yourself: ABE Window & Bystander Specificity Workbench
Adjust the target adenine protospacer position and TadA window sharpness (1 / σ_W) to optimize clean wild-type allele restoration while minimizing bystander deamination.
P_edit(x_i) = η_R-loop · exp(−(x_i − x_0)² / 2σ_W²) · k_deam / (k_deam + k_off)
Normalized Sensitivity (Stages 1–4)
Governing law: P_edit(x_i) = η_R-loop · exp(−(x_i − x_0)² / 2σ_W²) · k_deam / (k_deam + k_off)
Want to test this mechanism inside the full multi-module pillar observatory? Open the coupled interactive workbench on the Pillar Hub.
Protospacer Window Activity · Mode: Biophysical Sandbox · How to explore this model
- Test the four physical regimes: Click the four scenario presets above (1 · PAM Recognition & R-Loop through 4 · Bystander Window Control) to jump directly between experimental conditions.
- Perturb the primary control: Drag Target Adenine Protospacer Position (x_target) (3.0 to 9.0 pos (PAM=21)) and observe how the visual schematic and readouts respond.
- Explore coupled co-perturbation: Adjust Narrow-Window Deaminase Specificity (1 / σ_W) (0.6 to 2.0window sharpness) to test synergy against
P_edit(x_i) = η_R-loop · exp(−(x_i − x_0)² / 2σ_W²) · k_deam / (k_deam + k_off).
Many severe stop variants (such as W1282X or R553X) happen because a single G in DNA mutated into an A. Adenine base editing parks a guided enzyme over that 20-letter stretch of DNA and chemically converts the A back into a G-like letter—without cutting both strands of the DNA helix. Try adjusting the deaminase window width below: the art is making the spotlight narrow enough to fix the disease-causing A at position 6 without accidentally editing a neighboring 'bystander' A right next to it.
Questions to discuss with a CF care team: What is the difference between transient RNA therapy and durable DNA base editing in terms of safety testing and delivery?
What this diagram leaves out: Shows sequence-window specificity on naked target DNA; in vivo editing also depends on chromatin packing and airway basal cell reach.
Act IV · Clinical & Therapeutic Horizon
What ABE Can Edit and Where Prime Editing Takes
In human CF airway epithelia and intestinal organoids (Umbach et al., Sci. Transl. Med. 2026, PMID 42018671), ABE delivered as mRNA or viral vectors rescues chloride transport across W1282X, R553X, and canonical splice-acceptor 1717-1G>A—distinct from deep-intronic cryptic donors such as 3849+10kb C->T modeled in `aso_splice_model.py`, where inhaled antisense SPL84 is listed under NCT06429176 (ClinicalTrials.gov does not evaluate safety or efficacy; listing on the registry does not denote government endorsement.). Because base editors perform only transition substitutions (A→G or C→T), they cannot insert the three missing nucleotides of F508del or fix transversion stop codons like G542X (GGA → UGA), leading directly to Essay G.06: Prime Editing. Grounded in Lab Discovery Catalog modules: `base_editing_cftr`, `round179_abe8e_splice_base_editing_bridge`, `aso_splice_model`.
Translational boundaries for adenine base editing are established by the chemical scope of transition mutagenesis and protospacer targeting geometry. Base editors are strictly limited to transition mutations: adenine base editors convert A·T base pairs to G·C pairs, whereas cytosine base editors convert C·G pairs to T·A pairs. In CFTR biology, this transition chemistry provides direct restorative correction for historical G-to-A transition variants, most notably the canonical intron 11 splice-site mutation 1717-1G>A, and nonsense codons such as W1282X (UGG → UGA on mRNA, requiring an A-to-G reversion on the template strand) and R553X (CGA → UGA). Because base editing alters the chromosome itself, successful editing durably restores endogenous transcriptional regulation without requiring repeat dosing.
Conversely, base editors cannot rescue transversion mutations or multi-nucleotide structural indels. For example, G542X involves a GGA-to-UGA transversion that requires an A-to-C or U-to-G alteration; attempting to edit G542X with an adenine base editor deaminates the template strand to generate a CGA arginine codon (G542R) rather than authentic GGA glycine. Similarly, the three-base deletion F508del lies outside the reach of transition deaminases. For these complex structural alleles, prime editing or targeted gene insertion is required. Furthermore, safe clinical translation in airways necessitates transient delivery of base editor components via non-viral mRNA and chemically stabilized single-guide RNAs within lipid nanoparticles. This transient exposure restricts active deamination to a 48-hour window, suppressing transcriptome-wide off-target deamination while providing durable restorative conductance in patient-derived respiratory organoids.
Primary Anchor: Umbach et al. (2026), Science Translational Medicine 18(846):eadw8886 — Functional correction of the untreatable CFTR 1717-1G>A mutation through mRNA- and sgRNA-optimized base editing (PMID 42018671). (DOI 10.1126/scitranslmed.adw8886 · PubMed 42018671).
Governed by biophysical kernel src/scripts/ptc-abe-simulator.js and cross-checked in the Open Adenine Base Editing (ABE) Workbench on /cf/ptc/.
Resumen Científico en Español (ES-MX) · Folio G.05
Edición de Bases de Adenina (ABE8e / TadA): Conversión A·T a G·C Sin Cortes de Doble Cadena: Cómo la nickasa Cas9 fusionada con la desaminasa evolucionada TadA convierte codones de paro prematuro y sitios de empalme (1717-1G>A) dentro de una ventana de 5 nucleótidos.
Muchas variantes graves de Clase I—como W1282X, R553X, R1162X y 1717-1G>A—son transiciones G·C a A·T. La edición de bases de adenina (ABE) revierte exactamente ese cambio convirtiendo la adenina en inosina (leída como guanina) sin cortar la doble hélice de ADN.
En Palabras Sencillas: Muchas alteraciones genéticas graves que inhabilitan el canal de cloruro humano involucran una sola letra química errónea en el plano genético fundamental. En estos casos, una transición histórica accidental convirtió una instrucción original en una señal de paro prematuro o destruyó una unión indispensable para el empalme. Las herramientas tradicionales de edición genética intentan corregir cortando por completo ambas cadenas de la doble hélice. Sin embargo, debido a que las células adultas de las vías respiratorias rara vez se dividen, tienen grandes dificultades para reparar roturas de doble cadena sin introducir inserciones y deleciones destructivas. Los editores de bases eliminan este riesgo por completo. En lugar de cortar la hélice, una enzima de búsqueda abre una pequeña ventana de diez letras a lo largo de una sola cadena de ADN. Otra enzima acoplada reorganiza los átomos de la letra defectuosa, convirtiéndola en un compuesto que la célula lee como guanina. La información genética queda restaurada duraderamente sin romper el cromosoma.
Analogía Cotidiana: Considere a un restaurador experto que trabaja sobre un manuscrito en pergamino invaluable donde un error ortográfico altera el sentido legal de un decreto antiguo. Un reparador descuidado cortaría el párrafo completo con un bisturí, dejando un hueco que requeriría un parche tosco. En cambio, el conservador aplica una microgota de reactivo químico directamente sobre la tinta de la letra equivocada. La reacción modifica selectivamente el trazo erróneo convirtiéndolo en el carácter adecuado sin debilitar las fibras del pergamino. Un editor de bases de adenina trabaja en la cromatina con esa misma delicadeza química. Desenrolla una pequeña ventana de cinco letras de ADN y reorganiza los átomos de una sola base diana. La estructura del cromosoma permanece íntegra, evitando los daños estructurales de las roturas de doble cadena. No obstante, esta analogía deja de ser exacta porque la cromatina viva experimenta procesos activos de reparación de discordancias que no existen en el pergamino inerte.
Guía de Controles Interactivos: Antes de revisar las ecuaciones cinéticas que rigen la desaminación, el laboratorio interactivo le permite explorar las restricciones espaciales del complejo de edición de bases. El primer deslizador fija la posición del nucleótido diana dentro del espaciador, determinando en qué punto de la burbuja de ADN monocatenario se sitúa la base mutante respecto a la enzima. La eficiencia de edición alcanza su máximo cuando la base se alinea en la ventana óptima entre las posiciones cuatro y ocho. El segundo control regula la especificidad del editor en ventanas estrechas, ajustando la nitidez espacial del dominio catalítico. Observe las lecturas. Agudizar la ventana de acción eleva el rendimiento de edición deseada por encima del cincuenta por ciento, mientras reduce la desaminación de adeninas vecinas a menos del dos por ciento. Crucialmente, la tasa de roturas de doble cadena permanece en niveles despreciables menores al uno por ciento, permitiendo que la conductancia transepitelial de cloruro recupere más del cuarenta y cinco por ciento de sus valores normales.
Riesgo Biofísico y Clínico: Los denominadores nacionales no combinados contextualizan la edición de bases sin sumar cohortes entre jurisdicciones ni equiparar la inelegibilidad genotípica con eficacia clínica demostrada: CFFPR 2024 registra 92.3 % elegibles por genotipo, 26,277 de 29,314 (89.6 %) individuos elegibles con prescripción (3,037 o 10.4 % sin prescripción) y 8.0 % no elegibles; CCFR 2024 reporta 7.6 % no elegibles por edad y genotipo, 3,087 individuos en ETI (1,004 niños, 2,083 adultos) y 780 sin prescripción; UKCFR 2024 documenta 11,381 individuos totales (93 % de cobertura anual) y 8,830 asistentes a revisión en moduladores (80.3 % Kaftrio; 12.4 % sin uso previo); y ACFDR 2025 (31 dic 2025, sin trasplante) reporta 5.8 % no elegibles en adultos (2,326 con elegibilidad conocida) y 11.5 % (170 niños) en la cohorte pediátrica (1,473 niños). Mientras que 3849+10kb C->T permite bloqueo estérico con ASO (`aso_splice_model.py`, SPL84 en NCT06429176; ClinicalTrials.gov does not evaluate safety or efficacy; listing on the registry does not denote government endorsement.), la variante canónica 1717-1G>A exige reparación A·T a G·C demostrada por Umbach en 2026.
Derivación Formal: La probabilidad espacial y cinética de la edición de bases de adenina en la hebra de ADN monocatenario desplazada se formaliza mediante P_edit(x_i) = η_R-loop · exp(−(x_i − x_0)² / 2σ_W²) · k_deam / (k_deam + k_off). En esta expresión, η_R-loop describe la eficiencia termodinámica de formación del bucle R de Cas9, regida por la estabilidad del híbrido ARN guía–ADN frente a la energía de apertura de la doble hélice genómica. El factor gaussiano exp(−(x_i − x_0)² / 2σ_W²) modela el alcance estérico del conector peptídico flexible que une la desaminasa TadA al extremo N-terminal de Cas9, con posición media x_0 ≈ 5.5 y varianza de ventana σ_W². Dentro de esta burbuja expuesta, la probabilidad de que el nucleótido x_i se desamine depende de la competencia cinética entre la velocidad intrínseca de desaminación k_deam y la constante de disociación microscópica del bucle R k_off. Una vez que la adenosina diana se desamina a inosina, la resolución de la discordancia determina el resultado genómico duradero. El dominio nickasa D10A de Cas9 corta la cadena de ADN no editada, generando una mella específica tres nucleótidos antes del motivo PAM. Esta mella recluta factores endógenos de reparación de discordancias (MutSα/MutLα) y la exonucleasa 1 (EXO1), degradando la hebra no modificada y utilizando la hebra con inosina como molde. La ADN polimerasa δ rellena la brecha incorporando citosina frente a la inosina, fijando un intermediario I·C que se convierte en un par G·C duradero tras la replicación. En el simulador (src/scripts/ptc-abe-simulator.js), variantes TadA de ventana estrecha (como ABE8e-V106W) reducen σ_W, limitando la edición a la posición diana 6 y suprimiendo la desaminación de adeninas vecinas por debajo del 2 %.
Frontera Traslacional: Los límites traslacionales de la edición de bases de adenina están fijados por el alcance químico de las transiciones y la geometría del espaciador. Los editores de bases se restringen estrictamente a transiciones: los editores de adenina convierten pares A·T en G·C, mientras que los de citosina convierten C·G en T·A. En la biología de CFTR, esta química ofrece corrección terapéutica directa para variantes derivadas de transiciones históricas de G a A, como la mutación del intrón 11 1717-1G>A y los codones sin sentido W1282X (UGG → UGA en ARNm, requiriendo reversión A a G en la hebra molde) y R553X (CGA → UGA). Al modificar el cromosoma, la edición exitosa restablece de forma duradera la regulación transcripcional endógena sin requerir aplicaciones repetidas. Por el contrario, los editores de bases no pueden corregir transversiones ni deleciones de múltiples nucleótidos. Por ejemplo, G542X implica una transversión GGA a UGA que requiere modificar A a C o U a G; intentar editarla con ABE desamina la hebra molde produciendo un codón de arginina CGA (G542R) en lugar de glicina GGA. De igual forma, la deleción F508del escapa al alcance de las desaminasas de transición. Estas alteraciones complejas requieren edición de calidad o inserción génica. Asimismo, la seguridad clínica en vías respiratorias exige la entrega transitoria de los componentes del editor mediante ARNm no viral y ARN guía estabilizados en nanopartículas lipídicas, limitando la actividad a cuarenta y ocho horas y reduciendo cortes inespecíficos en organoides derivados de pacientes.
Historial de ediciones (REV 1.3): REV 1.0 (2026-11-01): Publicación inicial: ventana gaussiana en bucle R de Cas9 D10A y cinética de TadA-8e para 1717-1G>A y W1282X. · REV 1.1 (2026-11-04): Se incorporaron cotas de discriminación de bases vecinas y destilación bilingüe. · REV 1.3 (2026-11-10): Se integraron denominadores nacionales no combinados (CFFPR 2024, CCFR 2024, UKCFR 2024, ACFDR 2025) y el contraste frente a aso_splice_model.py y NCT06429176 (SPL84) en 3849+10kb C->T.
Horizonte terapéutico: Repara directamente el ADN cromosómico para que todo ARNm transcrito posteriormente sea nativo, evitando roturas de doble cadena (DSB) e inserciones/deleciones. Investigación independiente, no asesoramiento médico.
Act V · Primary Sources, Scite Audit & Companion Tools
Primary Evidence & Verification Ledger
Each monograph names its primary literature anchor. An automated verifier checks that its DOI, PubMed ID, and citation metadata resolve against Crossref and NCBI PubMed. Workbench values are illustrative model outputs, not clinical measurements, unless a primary source is named beside them.
- Primary Literature Receipt: Umbach et al. (2026), Science Translational Medicine 18(846):eadw8886 — Functional correction of the untreatable CFTR 1717-1G>A mutation through mRNA- and sgRNA-optimized base editing (PMID 42018671). (DOI 10.1126/scitranslmed.adw8886 · PubMed 42018671).
- Citation Context (Scite): Inspect how later literature cites 10.1126/scitranslmed.adw8886 on Scite.ai ·
Crossref + Scite verified (data/folio-citations.cache.json · fetched 2026-11-08 · 0 citing · 0 supporting · 0 mentioning · 0 contrasting). - Open-Source Computational & Epistemic Engine: OpenAlex, Unpaywall & cf-evidence-loop Splice-Variant Paper-Trail (DOI 10.7717/peerj.4375 · Scite Report).
- Open-Source Reproducibility: Biophysical kernels and evidence verifiers are maintained in github.com/senseiewok/cf-research and github.com/senseiewok/cf-skills.
For builders: method, companion SKILL.md, Python verifier, and edition history (REV 1.3)
cftr-adenine-base-editing-abe-SKILL.md · Kernel src/scripts/ptc-abe-simulator.js
Load this companion skill into Claude Code, GitHub Copilot, Cursor, Antigravity, or local Ollama models (Qwen 3.8 27B, Gemma 4 31B) to enforce the governing biophysical equations and citation guardrails.
View raw SKILL.md source
--- name: cftr-adenine-base-editing-abe description: AI-agnostic gene-editing skill for Adenine Base Editing (ABE8e/TadA), R-loop protospacer window geometry (positions 4-8), and bystander discrimination in CFTR. license: CC-BY-4.0 --- # G.05 — Adenine Base Editing (ABE8e / TadA) Without DNA Breaks ## Governing Biophysical / Mathematical Law P_edit(x_i) = η_R-loop · exp(−(x_i − x_0)² / 2σ_W²) · k_deam / (k_deam + k_off) ## Primary Evidence Receipt - **Citation:** Umbach et al. (2026), Science Translational Medicine 18(846):eadw8886 — Functional correction of the untreatable CFTR 1717-1G>A mutation through mRNA- and sgRNA-optimized base editing (PMID 42018671). - **DOI:** https://doi.org/10.1126/scitranslmed.adw8886 - **PubMed PMID:** https://pubmed.ncbi.nlm.nih.gov/42018671/ - **Answer Tier:** T2 · Sourced CRISPR Base-Editing & Kinetic Model ## Verification Rules - Verify strand polarity: ABE converts A*T to G*C (on coding strand A->G, or on template strand T->C on coding strand). - Evaluate bystander adenines within the protospacer 4-8 window (especially at W1282X and G542X). - Ground human airway CFTR base editing in Umbach et al. 2026 (DOI 10.1126/scitranslmed.adw8886, PMID 42018671). ## Safety & Clinical Boundary Research synthesis and biophysical modeling only — never medical advice, clinical guidance, or a treatment promise.
Verified 5-Act magazine flow, untruncated headings, and bilingual EN / ES-MX structural parity.
Audited clinical dignity boundaries, non-promise guardrails, and discovery catalog coverage.
check_folio_citations.py
Verified DOI 10.1126/scitranslmed.adw8886 against Crossref and NCBI PubMed metadata.
src/scripts/ptc-abe-simulator.js
Coupled governing relation P_edit(x_i) = η_R-loop · exp(−(x_i − x_0)² / to interactive Plates A & B.