A newly developed statistical framework called GapCNV improves the detection of copy number variants in low-input DNA sequencing data from both haploid and diploid genomes. Copy number variants are prevalent across diverse genomes, where they contribute directly to genome evolution and phenotypic diversity. However, identifying these alterations in low-cell and single-cell sequencing remains difficult due to amplification bias as well as shallow, uneven read counts across sequencing targets.
A major obstacle in low-input sequencing analysis involves selecting reference samples that establish a reliable baseline for defining copy-number gains or losses. Standard normalization approaches that use reference samples harboring common copy number variants can inadvertently cancel out authentic variant signals or generate false variant calls. GapCNV addresses this challenge by applying preliminary cell clustering alongside shared variant profiles to construct bin-specific pseudo-references derived specifically from cells inferred to occupy a copy-neutral state.
By generating these adaptive baselines, the framework limits the incorporation of common copy number variants into the normalization baseline, thereby preserving true shared variant signals. Researchers evaluated the framework across synthetic simulations run both with and without amplification biases, as well as biological datasets spanning Plasmodium falciparum, melanoma cell lines, and human breast tumors. In simulation tests, the framework demonstrated high precision while sustaining competitive sensitivity, showing distinct advantages for capturing short and common copy number variants.
When applied to Plasmodium falciparum data, the framework successfully detected known copy number variants and recovered short variants more frequently than benchmarked methods. Furthermore, analyses of human breast tumors and melanoma cell lines revealed that the method recovered bulk or single-cell derived copy number profiles across individual cells. These experimental outcomes indicate that constructing adaptive, bin-specific pseudo-references substantially enhances variant identification when reliable normal control samples are unavailable.
Subject of Research: Statistical framework for copy number variant detection in low-input sequencing
Article Title: GapCNV: A comprehensive framework for CNV detection in low-input DNA sequencing data
Article References: Yu, X., Qin, F., Liu, S., Brown, N. J., Fan, X., Lu, Q., Cai, G., Guler, J. L., & Xiao, F. (2026). GapCNV: A comprehensive framework for CNV detection in low-input DNA sequencing data. PLOS Computational Biology, 22(11), e1014959. https://doi.org/10.1371/journal.pcbi.1014959
Image Credits: AI Generated
DOI: 10.1371/journal.pcbi.1014959
Keywords: GapCNV, copy number variants, low-input DNA sequencing, single-cell sequencing, low-cell sequencing, pseudo-reference construction, copy-neutral state, cell clustering, normalization baseline, diploid genomes, haploid genomes, amplification bias, Plasmodium falciparum, human breast cancer, melanoma cell line, genome evolution, phenotypic diversity
Cite Scienmag News
Juliet Wilcox. (November 18, 2026). Adaptive Pseudo-References Improve Copy Number Variant Detection in Low-Input Sequencing. Scienmag. https://scienmag.com/adaptive-pseudo-references-improve-copy-number-variant-detection-in-low-input-sequencing/
Juliet Wilcox. "Adaptive Pseudo-References Improve Copy Number Variant Detection in Low-Input Sequencing." Scienmag, 18 November 2026, https://scienmag.com/adaptive-pseudo-references-improve-copy-number-variant-detection-in-low-input-sequencing/. Accessed 18 November 2026.
Juliet Wilcox. "Adaptive Pseudo-References Improve Copy Number Variant Detection in Low-Input Sequencing." Scienmag. November 18, 2026. https://scienmag.com/adaptive-pseudo-references-improve-copy-number-variant-detection-in-low-input-sequencing/








