generate_complex_data

Generates synthetic high-dimensional data with clusters and trends for testing
dimensionality reduction methods. Creates data with specified properties:<ul>
<li>Multiple clusters along a trend line</li>
<li>Variable density regions</li>
<li>Controllable noise levels</li>
<li>Optional visualization</li>
</ul>The function generates cluster centers along a trend line, adds points around those
centers with specified spread, and incorporates random noise to create high and
low density areas. The data is useful for testing dimensionality reduction and
visualization methods.

An implementation of the TopoLow algorithm, a novel, physics-inspired method for
antigenic cartography. TopoLow addresses significant challenges in mapping antigenic relationships,
especially from sparse and noisy experimental data. The package transforms cross-reactivity and
binding affinity measurements into accurate spatial representations in a phenotype space.
Key features include:
* Robust Mapping from Sparse Data: Effectively creates complete and consistent maps even with high
proportions of missing data (e.g., >95%).
* Physics-Inspired Optimization: Models antigens as particles connected by springs (for measured
interactions) and subject to repulsive forces (for missing interactions), reducing the need for
complex gradient computations.
* Automatic Dimensionality Detection: Employs a likelihood-based approach to determine the optimal
number of dimensions for the antigenic map, avoiding distortions common in methods with fixed low dimensions.
* Noise and Bias Reduction: Naturally mitigates experimental noise and bias through its network-based,
error-dampening mechanism.
* Antigenic Velocity Calculation: Introduces and quantifies "antigenic velocity," a vector that describes
the rate and direction of antigenic drift for each pathogen isolate. This can help identify cluster
transitions and potential lineage replacements.
* Broad Applicability: Analyzes data from various pathogens, including influenza, HIV, and Dengue viruses.
It can be applied to any continuous and relational phenotype under directional selection pressure.
Methods are described in Arhami and Rohani (2025) <doi:10.1093/bioinformatics/btaf372>.

Omid Arhami

topolow

Antigenic Mapping and Antigenic Velocity Algorithm

generate_complex_data function

<dl><dt>n_points</dt>
<dd>Integer number of points to generate</dd>
<dt>n_dim</dt>
<dd>Integer number of dimensions</dd>
<dt>n_clusters</dt>
<dd>Integer number of clusters</dd>
<dt>cluster_spread</dt>
<dd>Numeric controlling cluster variance</dd>
<dt>fig_name</dt>
<dd>Character path to save visualization (optional)</dd></dl>

Arguments

Generates synthetic high-dimensional data with clusters and trends for testing
dimensionality reduction methods. Creates data with specified properties:<ul>
<li>Multiple clusters along a trend line</li>
<li>Variable density regions</li>
<li>Controllable noise levels</li>
<li>Optional visualization</li>
</ul>

The function generates cluster centers along a trend line, adds points around those
centers with specified spread, and incorporates random noise to create high and
low density areas. The data is useful for testing dimensionality reduction and
visualization methods.

Generate Complex High-Dimensional Data for Testing — generate_complex_data

<dl>

<dt>n_points</dt>
<dd>Integer number of points to generate</dd>


<dt>n_dim</dt>
<dd>Integer number of dimensions</dd>


<dt>n_clusters</dt>
<dd>Integer number of clusters</dd>


<dt>cluster_spread</dt>
<dd>Numeric controlling cluster variance</dd>


<dt>fig_name</dt>
<dd>Character path to save visualization (optional)</dd>

</dl>

generate_complex_data: Generate Complex High-Dimensional Data for Testing

Description

Usage

Value

Arguments

Examples