Eculizumab and pegcetacoplan both inhibit complement — but they act at different nodes. When a patient cohort is activated predominantly at C3, a C5 inhibitor may leave most of the damage unblocked. Transcriptomics lets you look upstream and answer the stratification question before a trial begins.

Why Pathway Position Matters in Complement Therapeutics

The complement system is not a single target; it is a proteolytic cascade with distinct branch points, each susceptible to a different drug class. Eculizumab and ravulizumab (anti-C5 monoclonals) block the cleavage of C5 into C5a and C5b, preventing membrane attack complex (MAC) formation downstream. Pegcetacoplan (anti-C3 peptide) acts further upstream, blocking amplification at C3 convertase and reducing both MAC formation and the generation of C3a and C3b opsonins.

The therapeutic implication is precise: in a cohort where disease pathology is driven by excessive C3 opsonization — as seen in some peripheral neuropathies — anti-C5 therapy leaves the upstream C3 arm intact and may achieve only partial suppression. Conversely, in a MAC-driven membranolytic disease, C5 inhibition is directly on-mechanism. Choosing the right node requires knowing where, in a given tissue and patient population, the cascade is most active.

"In CIDP, IgG4 autoantibodies against nodal and paranodal proteins activate complement at the axonal surface. The dominant damage mechanism — C3d opsonization versus MAC pore formation — varies by antigen and biopsy location."

Bulk RNA-seq from blood or nerve biopsy offers a practical window into this question. Complement system genes — C1QB, C3, C4A, C5, CFB, CFD, CFH, CFI — are robustly transcribed in immune infiltrates, Schwann cells, and macrophages. Their relative expression profiles, interpreted through pathway-level enrichment, can distinguish cohorts where the classical C3 arm dominates from those where late-pathway (C5/MAC) activation is the prevailing signature.

The Three-Step Chained Workflow

BioMate executes this analysis as a sequential chain: each step's output feeds the next, and the final output is a ranked table of complement sub-pathway activation scores with drug-class implications attached.

Reference Datasets: CIDP and Myasthenia Gravis

Two GEO datasets serve as the validation benchmark for this workflow. GSE138426 provides nerve biopsy transcriptomics from CIDP patients versus healthy controls — a tissue-proximal view of complement activation at the site of injury, where macrophage infiltration and Schwann cell reactivity amplify both classical and alternative pathway signals. GSE51867 provides PBMC gene expression from myasthenia gravis patients, capturing the peripheral blood immune milieu including classical pathway activation driven by AChR-specific IgG1 antibodies.

MG with anti-AChR antibodies is the best-characterized complement-mediated neuromuscular disease: IgG1 autoantibodies fix C1q at the neuromuscular junction, activating the classical pathway and driving MAC-mediated destruction of post-synaptic folds. This pathway architecture — classical, C5-convergent — is why eculizumab achieved phase III efficacy in refractory generalized MG (REGAIN trial, Hehir et al., 2017). CIDP presents a more heterogeneous picture: nodal IgG4 antibodies do not fix complement, but sural nerve biopsies show C3d deposition on Schwann cells and axons, implicating macrophage-mediated alternative pathway amplification upstream of C5.

Heatmap: Complement Gene Expression Across Patient Samples

Before sub-pathway ranking, it is useful to visualize the raw normalized expression of key complement genes across individual samples. The heatmap below represents variance-stabilizing-transformed (VST) counts for eight complement-related genes across ten CIDP samples and five controls, revealing inter-sample heterogeneity that the aggregate enrichment score would otherwise collapse.

Extending to GBS: Radar Chart of Cohort-Level Sub-Pathway Scores

Adding Guillain-Barré syndrome (GBS) as a third comparator cohort illustrates why the sub-pathway decomposition is clinically valuable. GBS — in particular the axonal variants AMAN and AMSAN — features anti-ganglioside antibodies that fix complement at nodes of Ranvier and drive MAC-mediated axonal injury. Published nerve biopsy data (Hafer-Macko et al.) demonstrate intense C3d and C5b-9 co-deposition in acute GBS axons. The cohort therefore provides a contrast case to CIDP where C5-convertase and terminal pathway scores should be disproportionately elevated.

Drug-Class Implications from the Dominant Index

The final output of the BioMate chain is a per-cohort dominant-pathway index with drug-class annotations. The table below summarizes the expected stratification across the three diseases, with therapeutics mapped to their mechanism of action and the enrichment signature that predicts on-target activity.

Cemdisiran: siRNA Silencing as a Next-Generation C5 Strategy

Cemdisiran (Alnylam Pharmaceuticals) is a GalNAc-conjugated siRNA that silences hepatic C5 gene expression, reducing circulating C5 protein and thereby suppressing MAC formation without occupying the C5 protein target. Unlike monoclonal antibodies, which require dosing in proportion to circulating C5 concentration, cemdisiran ablates the supply of the substrate protein itself. This makes it particularly attractive for diseases where C5 production is amplified — including acute-phase inflammatory states — and for combination strategies with eculizumab, where cemdisiran pre-dosing reduces the stoichiometric burden on the antibody.

The transcriptomic stratification workflow described here is directly applicable to next-indication selection for cemdisiran: if a patient cohort presents a C5-dominant enrichment profile (high MAC/Terminal GSEA score, elevated C5 and C6–C9 expression relative to CFB and C3), the mechanistic rationale for C5 silencing is clear. If the profile is C3-dominant, the upstream protein supply node matters less than the convertase activity itself, and the siRNA strategy may be less effective on its own.