Gene expression tests for non-coding variants compared in neurodevelopmental disorders

Non-coding variants linked to neurodevelopmental disorders are notoriously hard to interpret. Unlike mutations that disrupt a protein directly, changes in regulatory DNA often exert subtle effects that depend on the specific epigenetic state of a locus at a particular developmental stage. A new study in Genome Medicine, published as an open-access research article on 27 October 2026, systematically evaluated two transcriptomic strategies for detecting when such variants actually matter, and found that the choice between them is less obvious than many researchers assume.

The most reliable sign that a non-coding variant is functionally relevant is transcriptional mis-regulation of nearby genes. RNA-sequencing studies usually hunt for this signal using differential gene expression analysis, which compares variant carriers with healthy controls. That approach has a practical drawback: it typically requires expression changes of more than twofold to reach statistical significance, and it depends on the availability of matched control samples. An alternative is allele-specific expression analysis, which compares the activity of the two alleles within a single heterozygous individual — the intact chromosome against the one carrying the variant. Because both alleles share the same cellular environment, ASE can in principle detect much subtler regulatory effects.

To find out which strategy performs better in practice, a team led by Veniamin Fishman combined transcriptome profiling of induced pluripotent stem cells and neural progenitor cells with in silico modeling of gene-expression changes to estimate statistical power. They directly compared the two designs: differential expression between carriers and non-carriers, and allele-specific expression within heterozygous individuals. The simulations and empirical data pointed in a direction that challenges the intuitive appeal of within-sample comparisons. Differential expression had higher power for most genes, and it retained that advantage even when the researchers modeled substantial inter-individual or biological variability.

Allele-specific expression outperformed differential expression only under narrower conditions: for genes that carried a heterozygous single-nucleotide variant and were sequenced at high depth. In other words, the sensitivity advantage of ASE is real but conditional, and it comes with a requirement for sequencing depth that not every study can afford. The authors concluded that the two assays are complementary rather than competing strategies, and that the right choice depends on the gene, the variant and the resources available.

The study also included a clinical test case that illustrates both the promise and the limits of these methods. The team studied a patient carrying a translocation breakpoint near AUTS2, a gene implicated in neurodevelopmental disease. To enable sensitive allele-specific profiling, they used CRISPR-Cas9 base editing to introduce a synonymous single-nucleotide variant adjacent to the candidate regulatory variant in patient-derived neural progenitor cells. This synonymous change acts as a harmless molecular barcode, labeling the affected AUTS2 allele so its expression can be tracked separately from the intact allele. They generated iPSC and NPC lines carrying the labeled alleles and devised an inexpensive target-enrichment assay for high-sensitivity ASE profiling, workable both in whole-transcriptome and targeted RNA sequencing.

Despite this technical achievement, the biological answer remained elusive. Neither the whole-transcriptome approach nor the targeted assay detected gene misregulation consistent with the patient’s phenotype under current biological models.

The practical payoff of the work is a decision framework. Researchers validating non-coding variants implicated in neurodevelopmental disorders can now weigh the trade-offs between differential expression designs, which need controls but offer broad power, and allele-specific designs, which need no controls and can detect subtle effects but only when a heterozygous variant can be labeled or exploited and read out at high depth.

Subject of Research: Comparing differential and allele-specific gene expression approaches for interpreting non-coding variants in neurodevelopmental disorders

Article Title: Evaluating gene expression-based approaches for functional assessment of non-coding variants in neurodevelopmental disorders

Article References: Khabarova, A., Yan, A., Salnikov, P., Torgunakov, N., Tarkhova, A., Gridina, M., Musatova, E., Oliveira, D., de Castro, M. V., Zatz, M., da Rocha, L. A., Bertola, D. R., da Silva, M. V. R., Nurislamov, A., Koksharova, G., & Fishman, V. (2026). Evaluating gene expression-based approaches for functional assessment of non-coding variants in neurodevelopmental disorders. Genome Medicine. https://doi.org/10.1186/s13073-026-01820-w

Image Credits: AI Generated

DOI: 10.1186/s13073-026-01820-w

Keywords: non-coding variants, neurodevelopmental disorders, gene expression, RNA-seq, allele-specific expression, differential expression, CRISPR-Cas9, iPSCs, neural progenitor cells, AUTS2, statistical power, Genome Medicine