Skip to content

UC-5.1 — Sample-Compound Class Interaction

Module: 5 – Modeling Interactions of Samples, Genes, and Compounds
Visualization type: Chord diagram (bipartite sample–compound-class interaction network)
Primary inputs: BioRemPP results table with sample and compoundclass columns
Primary outputs: Interaction matrix of samples × compound classes (co-occurrence counts)

Scientific Question and Rationale

Question: Which samples are most co-annotated with which chemical classes, and what might this reveal about their annotation coverage patterns?

This use case maps the co-annotation frequency between each biological sample and the different classes of chemical compounds. By summarizing how often each sample is co-annotated with each compound class, and visualizing these relationships as a chord diagram, the analysis can provide an intuitive, systems-level overview of the co-annotation landscape. It may highlight samples with broad compound class coverage and those with narrowly focused co-annotation patterns.

Data and Inputs

  • Primary data source: BioRemPP_Results.xlsx or BioRemPP_Results.csv
  • Key columns:
  • sample – identifier for each biological sample
  • compoundclass – categorical label for the chemical class of each compound
  • Accepted format: semicolon-delimited text table (.txt or .csv)
  • Derived structure: interaction matrix with:
  • rows = samples
  • columns = compound classes
  • cell = interaction count for each sample–class pair

Analytical Workflow

  1. Data Loading
    The primary results table (BioRemPP_Results.xlsx or BioRemPP_Results.csv) is loaded from its semicolon-delimited format.

  2. Filtering
    The dataset is filtered to retain only complete entries containing both a valid sample and a compoundclass. Rows with missing values in either field are removed.

  3. Aggregation (Interaction Strength)
    The filtered data is grouped by each unique (sample, compoundclass) pair.

  4. For each pair, the total number of co-occurrences (rows) is counted.

  5. This count serves as the interaction strength between that sample and that compound class.

  6. Chord Matrix Construction
    The aggregated counts are arranged into a matrix or edge list suitable for chord diagram rendering, where:

  7. each sample is treated as a source node,
  8. each compound class is treated as a target node, and

  9. the edge weight is the interaction count.

  10. Rendering
    A chord diagram is generated:

  11. arcs on the circumference represent samples and compound classes,
  12. ribbons (chords) connect each sample to the classes with which it interacts,
  13. chord thickness encodes interaction strength.

How to Read the Plot

  • Outer Arcs (Nodes)
    Each colored arc along the circle represents either:
  • a Sample, or

  • a Compound Class.
    The length of an arc is proportional to the total number of interactions (sum of counts) associated with that entity.

  • Chords (Ribbons)
    The ribbons spanning between arcs represent individual Sample–Compound Class interactions:

  • one end of the ribbon is anchored at a sample arc,

  • the other end at a compound class arc.

  • Chord Thickness
    The thickness of a chord at its connection points is proportional to the interaction count for that sample–class pair:

  • thicker chords indicate stronger associations (more co-occurrences),
  • thinner chords indicate weaker or less frequent associations.

Representative Output

The image below illustrates a representative output generated by this use case using the example dataset.

Click on the image to enlarge and explore details.

Representative output for UC-5.1

Interpretation and Key Messages

  • Strong Co-annotation Associations A thick chord between a specific sample and a given compound class may indicate a strong co-annotation frequency:
  • the sample is co-annotated with many compounds from that class in the database,

  • which may suggest a concentrated annotation profile for that chemical domain, warranting further investigation.

  • Broad vs. Narrow Annotation Profiles

  • A sample with multiple substantial chords connecting to several compound classes shows broad compound class co-annotation coverage across diverse chemical spaces.

  • A sample whose representation is dominated by one or a few very thick chords shows narrower co-annotation coverage concentrated in specific chemical classes.

  • Highly Co-annotated Chemical Classes Compound classes that receive many thick chords from multiple samples may emerge as:

  • widely co-annotated classes in the dataset, and

  • potential focal points for further investigation, as they are broadly represented across the available samples.

  • Co-annotation Landscape Overview The overall pattern of chords can provide a global view of how co-annotations are distributed:

  • identifying central samples and classes that structure the network,
  • potentially revealing imbalances, redundancies, or gaps in annotation coverage.

Reproducibility and Assumptions

  • Input Format
    The analysis assumes a semicolon-delimited table containing at least the columns sample and compoundclass.

  • Interaction Definition
    Interaction strength is defined as the total number of co-occurrence records for each (sample, compoundclass) pair in the raw table:

  • multiple entries for the same pair (e.g., different compounds within the same class, or multiple genes linked to that class) increase the aggregate count,

  • the chord diagram thus reflects overall interaction intensity rather than unique compound counts.

  • Scope of Interpretation The chord diagram summarizes co-annotation frequency, not mechanistic detail:

  • it does not directly encode pathway completeness, kinetic efficiency, or regulatory control,
  • but it can provide a high-level annotation map that may guide more detailed, pathway-focused or gene-level analyses in subsequent use cases.

Activity diagram of the use case

Click on the image to enlarge and explore details.

Activity diagram of the use case