Molecular AI & Methods / 2026

HeFS.

Helper-enhanced feature selection via Pareto-optimized genetic search.

Journal of Advanced Research · advance online, 2026

Richard’s author role: Co-corresponding author

Read paper ↗Official code & data resources ↗

01 / The problem

A feature subset can miss complementary information left among the features discarded by an existing method.

02 / The core idea

Find complementary helper sets with biased initialization, proportion-guided mutation and Pareto-optimized genetic search.

How the idea flows

FIGURE 01 / ORIGINAL CONCEPT SKETCH

Existing feature subset
Discarded feature space
Complementary helper candidates
Pareto genetic search
Evaluate enhanced subset

Original concept sketch; not a reproduced paper figure.

03 / What the public evidence shows

The final publication evaluates the method across 18 benchmark datasets. This site does not turn reported relative gains into percentage-point gains or invent a standalone GitHub repository.

Source: HeFS ↗

04 / Where the claim stops

Feature selection is a general method, not a peptide-specific result. The paper is advance-online/in press in 2026; no volume or page range is invented.

An editorial scope note, not an unverified quotation of author limitations.