Affinity purification–mass spectrometry (AP-MS) and network analysis sit on two sides of one file. AP-MS reports bait-centered co-enrichment after a defined capture and liquid chromatography–tandem mass spectrometry (LC-MS/MS). Network analysis draws that ranked list as a graph and can overlay public edges or gene-ontology terms. The drawing does not convert the list into a structured interaction model, and it does not explain how partners relate biologically.

Signaling pathways, complexes, and disease mechanisms remain contexts in which a bait may be tested. They are not outputs of the spectrometer or of the layout algorithm. A node is a protein name. An edge is a rule the analyst wrote: co-enrichment with this bait, shared recovery across related baits, or a labeled literature link. For what the underlying evidence is allowed to support, Understanding AP-MS: A Comprehensive Guide provides a complementary evidence page.

Researchers who already have a bait format and need a ranked co-enrichment table before any graph is drawn can also review the MtoZ Biolabs Affinity Purification-Mass Spectrometry Service for project-specific feasibility and analysis planning.

What AP-MS Hands the Graph

Typical output remains identified proteins in bait and control purifications, quantitative bait-to-control contrast, and a filtered candidate list after replicate review. That list is association under the purification used. It does not define topology, direct versus indirect binding, or functional coupling among recovered partners.

Network analysis is needed after AP-MS only in the narrow sense that a long table is hard to inspect. Organizing first-order neighbors around the bait, marking proteins recovered with more than one related bait, and overlaying a pathway term can make review explicit. It does not distinguish a coherent complex from background, shared contaminants, and secondary associations. A hub can be a scaffold. It can also be a sticky contaminant. Treating the list as input to an interactome model raises the claim without adding a measurement.

Figure 1. Left: bait-versus-control table. Right: bait-centered graph with co-enrichment edges. The layout is not a second experiment.

Filter the table before drawing edges. A dense figure is not a more complete biology.

Experiment-First Edges, Then a Labeled Overlay

The original layers stay, with stricter verbs.

Interactor filtering uses bait-versus-control contrast, replicate support, and a contaminant resource. Do not subtract every control identification. Rank specificity. Spectral counts are identification support. They are not abundance. For that read order, How to Analyze IP Mass Spectrometry Results? provides a complementary analysis-order page.

Edge definition names the rule. An AP-MS-only graph usually links the bait to each retained prey. Shared recovery across related baits can add a second experimental class. Public database edges are a third class. They must keep a visible label. Silent merge with experimental edges is how a literature neighbor becomes a copurifying partner.

Node annotation carries accession, condition, and the score that survived filtering. Abundance language is allowed only when the acquisition was quantitative and bait recovery was matched.

Functional mapping overlays gene ontology, Reactome, KEGG, or a complex catalog. Those terms annotate the set. They do not translate topology into signaling cascades or membrane trafficking. For what a public catalog actually stores, What Are the Protein Protein Interaction Databases? provides a complementary database page. For overlay versus a measured contact, How to Build and Analyze Protein-Protein Interaction Networks? provides a complementary build-versus-overlay page.

Weak filtering still becomes false hubs. Transparent edge criteria matter more than the visualization tool.

Figure 2. Filter, draw experimental edges, then overlay literature or ontology with a separate style. A hub can be a contaminant.

What the Original Goal Table Can Support

The original five goals remain usable if the network focus is a drawing rule, not a biological grant.

Study goal	What the graph may show	What it does not grant
Single bait plus controls	First-order neighbors around the bait	An interactome map
Tagged complex component	A dense cluster of co-recovered names	Subunit organization
Treatment versus vehicle	Nodes that change after matched capture	PPI remodeling
Multiple related baits	Shared experimental recovery	Pathway expansion
Drug-treated AP-MS	Gain or loss of co-enriched names	Target discovery

Reciprocal AP-MS, Co-IP, mutagenesis, or a pairwise binding assay remain later tests. They are not typical next steps that the figure completes.

Bait-centered discovery in one experiment can recover many names. That is not proteome-scale input. Low-abundance partners missed in the MS file stay off the graph. Completeness of the drawing follows detection depth, not the layout.

Reporting, Not a Deliverable Package

A report that is usable for a figure separates three edge classes: AP-MS-derived co-enrichment, literature-supported links, and confirmed direct interactions. Method text must name bait format, controls, filtering threshold, and the edge rule. A validation priority list that mixes centrality with gene-ontology enrichment is a hypothesis queue. It is not a confirmed target list.

Disease-mechanism, patient-relevant, and engagement-linked language sits outside the graph unless a separately validated assay is already in place.

Figure 3. Keep: bait-centered drawing and a labeled overlay. Not granted: topology, direct binding, hub-as-core-complex, or a drug target.

Experimental Limits

AP-MS plus network analysis reduces a project to one capture, one ranked list, and one set of drawing rules. Diagnostic interactome or target-nomination language sits outside ordinary research reporting.

Project Planning and Service Scope

A usable plan names the bait, the matched control, the edge rule, and whether any public overlay will be labeled, before the first lysate is bound. MtoZ Biolabs scopes AP-MS on Orbitrap-class LC-MS/MS. A graph, if requested, still follows those arms. It does not become an interactome model by being drawn.

Once the next sentence is a reciprocal or endogenous antibody pellet rather than a labeled figure, the project leaves this page. Researchers who need that pellet can review the MtoZ Biolabs Co-IP-MS Protein Interaction Analysis Service for project-specific sample evaluation, feasibility assessment, and workflow planning.

Frequently Asked Questions

1. What does network analysis add to AP-MS?

A labeled drawing of the filtered list, and an optional public overlay. It does not add a new measurement.

2. Does an AP-MS network prove direct protein-protein interactions?

No. Co-enrichment edges are associations under that purification. Direct binding is a later assay.

3. Can a network be built from a single replicate?

A sketch is possible. Single-run names become false hubs easily.

4. Should public PPI databases be included?

Only as a separately labeled overlay. Do not merge them with experimental edges.

5. Does a hub in the graph mean a core complex subunit?

No. Hubs are also chaperones and frequent bead binders.

Reference

1. Gingras A.C., Gstaiger M., Raught B., Aebersold R. (2007). Analysis of protein complexes using mass spectrometry. Nat. Rev. Mol. Cell Biol., 8, 645-654. https://doi.org/10.1038/nrm2208
2. Mellacheruvu D., Wright Z., Couzens A.L., et al. (2013). The CRAPome: a contaminant repository for affinity purification-mass spectrometry data. Nat. Methods, 10, 730-736. https://doi.org/10.1038/nmeth.2557
3. Choi H., Larsen B., Lin Z.Y., et al. (2011). SAINT: probabilistic scoring of affinity purification-mass spectrometry data. Nat. Methods, 8, 70-73. https://doi.org/10.1038/nmeth.1541
4. Huttlin E.L., Ting L., Bruckner R.J., et al. (2015). The BioPlex network: a systematic exploration of the human interactome. Cell, 162, 425-440. https://doi.org/10.1016/j.cell.2015.06.043
5. Szklarczyk D., Kirsch R., Koutrouli M., et al. (2023). The STRING database in 2023: protein-protein association networks and functional enrichment analyses for any sequenced genome of interest. Nucleic Acids Res., 51, D638-D646. https://doi.org/10.1093/nar/gkac1000