About the Atlas

The Atlas of Computational Cell Reprogramming is a curated resource dedicated to the emerging field of computational cell fate engineering.

A central challenge in cell reprogramming is identifying interventions that steer a starting cell population toward a desired identity or functional state. Pluripotency induction redirects a somatic cell toward a pluripotent state. Directed differentiation guides a pluripotent or progenitor cell along a selected developmental path toward a mature lineage. Transdifferentiation converts a cell of one mature lineage into another without a pluripotent intermediate. Phenotype reprogramming redirects a cell from an undesired to a desired functional state while preserving lineage identity. The four settings differ in the transition each seeks, but all pose the same question: which feasible perturbations move a starting population to a chosen target state?

Despite rapid progress, the field remains fragmented. Computational methods are scattered across disciplines, evaluated using different assumptions, and often difficult to compare directly.

The Atlas was created to address this challenge. Our goal is to provide a continuously maintained map of computational approaches for cell reprogramming and cellular intervention design. Each method is systematically annotated, classified, and linked to its original publication. Beyond cataloging methods, the Atlas seeks to identify broader patterns across the field, including methodological trends, biological assumptions, validation strategies, reproducibility practices, and emerging technological directions.

The Atlas is built around a unifying perspective: many cell reprogramming methods can be viewed as attempts to solve a common intervention-design problem, namely finding molecular perturbations that move a cell from an initial state toward a desired target state. While biological settings differ substantially, this shared computational perspective provides a framework for comparing methods that would otherwise appear unrelated. The formal treatment, explicitness classes (C0–C3), and setting definitions are on the framework page.

The initial release is built around the 60 peer-reviewed computational methods audited for the manuscript under review Computational Cell Reprogramming: A Systematic Review of Cellular Intervention Design (2013–2026). The live index currently lists 60 methods and continues to grow as new peer-reviewed intervention-design methods are curated in.

The index covers computational methods across pluripotency induction, directed differentiation, transdifferentiation, and phenotype reprogramming, and across application areas that include regenerative medicine, immunology, aging, and cancer reversion.

The Atlas is maintained by the Vera-Licona Research Group. Inclusion, classification, and updates follow the public gates on this page. Nominations are reviewed against those gates through Curated Submissions.

The Atlas is a continuously updated and curated resource for computational cell reprogramming. New method nominations, corrections, reproducibility updates, and community feedback are welcomed through the Atlas curation process; see Curated Submissions.

Data & methodology: M1–M3 inclusion gates, annotation schema, releases, and citation.

Data & methodology

Inclusion criteria

A peer-reviewed method enters the Atlas when three eligibility gates all pass. The same gates selected the 60 methods audited in the manuscript under review (Supporting Information S1). They are inverse-design conditions: each asks whether the paper specifies an intervention-design procedure. A published name, a user-facing artifact, and software availability are recorded in the reproducibility audit.

M1. Actionable intervention rule

The study specifies a computational decision procedure that identifies, ranks, combines, or controls actionable interventions (u). A forward or descriptive model passes only when it is coupled to an intervention-selection or control procedure. Class C3 methods often contain simulators. A simulator alone fails this gate.

M2. Cell-state-conditioned objective

The decision procedure is conditioned on a biologically defined cellular source, target, contrast, trajectory endpoint, attractor, or undesirable-versus-desired phenotype. Generic regulator-enrichment tools that take an untyped gene set fail this gate.

M3. Reusable computational method object

The intervention-selection procedure is specified well enough to apply to another eligible instance using data of the required type. Computation must be a substantive part of the paper's methodological contribution, beyond routine processing or post hoc characterization. The rule can be stated independently of a particular cloning, barcoding, or screening platform.

Integrated experimental and computational screens are decided under M3. Computation must connect perturbations to a specified target state through target matching, candidate nomination, combination optimization, or another explicit intervention-selection rule. Barcode counting, standard differential-expression analysis, quality control, or post hoc mechanistic analysis alone fails this gate.

The gates also exclude forward simulators that do not search, rank, or nominate an intervention; landscape reconstruction, trajectory inference, optimal transport, or gene-regulatory network inference without an intervention-selection step; reviews, perspectives, protocols, and purely experimental papers with no computational nomination method; and records that have not been peer reviewed.

Annotation schema

Each indexed method receives:

• One explicitness class placement (C0, C1, C2, or C3) with a written rationale tied to the role the intervention $u$ plays in the method's objective. A publication with two selection objects is listed under each class and kept out of the single-class counts.
• One representation family (signature or state matching, regulatory network influence, or executable intervention model).
• One or more settings (I, D, T, P), taken from the original publication.
• An audit of code availability, reproducibility (0 to 4), and FAIR4RS compliance (0 to 5), refreshed periodically.

Interactive audit summaries and the scoring rubric are on the reproducibility page; per-method scores appear on each method page.

Maintained index

The Atlas currently lists 60 methods and grows as new peer reviewed intervention design methods are added. Each release is tagged (v0.2.2 is current); the changelog records additions, replacements, and audit refreshes. The SI_2 table freezes the original 60 method audit (2026-10-19). Headline reproducibility statistics in the manuscript under review refer to that frozen set; methods added afterward are tagged on their method pages.

Every method page shows its last_audited date. Major releases are archived on Zenodo with a citable DOI (first deposit not yet made); this site remains the live catalog between deposits.

Manuscript under review

Vera-Licona P. Computational Cell Reprogramming: A Systematic Review of Cellular Intervention Design. Manuscript under review.

The manuscript under review introduces the four explicitness classes (C0, C1, C2, C3), the shared inverse design framing across reprogramming settings, and the reproducibility audit summarized on the reproducibility page.

BibTeX: manuscript under review

@misc{veralicona2026reprogramming,
  author  = {Vera-Licona, Paola},
  title   = {Computational Cell Reprogramming: A Systematic Review of Cellular Intervention Design},
  year    = {2026},
  note    = {Manuscript under review},
}

The manuscript PDF, DOI, and supplementary tables will be linked here when they are available.

Citation: this site

To cite the Atlas:

Vera-Licona, P. (2026). Atlas of Computational Cell Reprogramming, version v0.2.2.

A Zenodo DOI will be added here when the Atlas snapshot is deposited.