L1 · Network-informed inverse design heuristics I D T

CellNet

Cahan P, Li H, Morris SA, Lummertz da Rocha E, Daley GQ, Collins JJ

2014 · Cell

Canonical Level 1 method. Reconstructs cell-type-specific GRNs from expression data and prioritizes regulators whose perturbation is expected to restore the target network.

Abstract

From the original paper, Cell · PubMed

Somatic cell reprogramming, directed differentiation of pluripotent stem cells, and direct conversions between differentiated cell lineages represent powerful approaches to engineer cells for research and regenerative medicine. We have developed CellNet, a network biology platform that more accurately assesses the fidelity of cellular engineering than existing methodologies and generates hypotheses for improving cell derivations. Analyzing expression data from 56 published reports, we found that cells derived via directed differentiation more closely resemble their in vivo counterparts than products of direct conversion, as reflected by the establishment of target cell-type gene regulatory networks (GRNs). Furthermore, we discovered that directly converted cells fail to adequately silence expression programs of the starting population and that the establishment of unintended GRNs is common to virtually every cellular engineering paradigm. CellNet provides a platform for quantifying how closely engineered cell populations resemble their target cell type and a rational strategy to guide enhanced cellular engineering.

Summary

CellNet is a transcriptome-driven network method that reconstructs cell-type-specific gene regulatory networks (GRNs) from large compendia of expression data and uses those networks as references for assessing how closely engineered cells resemble their intended target identity. When used for reprogramming design, it prioritizes regulators whose perturbation is expected to restore the target network.

The argument is network-informed and source–target aware, but it is not a forward perturbation model: candidate $u$ is scored against influential network positions in the target GRN, not by simulating the post-intervention state.

Why this level

Level 1 because the method scores candidate interventions through static regulatory-network influence relative to a target GRN, without computing $\hat P_{S,u}$ . Representation family is regulatory-network influence: the method's central object is the cell-type-specific GRN. CellNet is the canonical Level 1 exemplar in §3.2 of the review.