Evolutionary and functional constraints structure human gene research visibility

Biomedical research effort is distributed highly unevenly across human genes, with a small subset dominating the scientific literature while thousands remain sparsely studied. Whether this imbalance reflects intrinsic biological importance or historically reinforced research bias remains unclear. Understanding how research attention relates to gene properties is essential for more systematic exploration of the human genome.

Here, we quantify gene-level publication patterns and integrate sequence features, evolutionary constraint, gene age, expression, and disease associations across stratified gene sets. Using standardized MANE Select annotations, we show that publication counts follow a strongly heavy-tailed distribution. Highly studied genes cluster within a narrow GC-content regime and exhibit lower nonsynonymous substitution rates and lower dN/dS ratios, consistent with stronger long-term evolutionary constraint. In contrast, genes sampled from below rank 10,000 are enriched for evolutionarily younger loci and display moderately elevated dN and dN/dS values, reduced expression magnitude, and increased tissue specificity. At the disease level, research attention concentrates within a limited number of dominant domains, particularly cancer, respiratory, and vascular diseases, whereas congenital and rare disease categories remain comparatively underrepresented. Genes associated with orphan diseases show significantly reduced publication counts.

Together, these results demonstrate that research attention is systematically structured across evolutionary, molecular, and disease dimensions. The least-studied genes represent a distinct and underexplored portion of the genome, characterized by features that may reduce experimental tractability. These findings highlight the need for bias-aware research prioritization strategies to broaden discovery and ensure more comprehensive characterization of human genes.