PHACTn enables training-free, context-independent inference of nucleotide variant tolerance across the genome

Abstract

Accurate prediction of single-nucleotide variant (SNV) tolerability across the entire human genome remains a fundamental challenge in computational genomics, particularly for non-coding regions where the regulatory landscape is vast and poorly understood. Machine learning classifiers suffer from data circularity and demographic bias, while genomic language models demand massive computational resources and offer little biological interpretability. Here, we present PHACTn (Phylogeny-Aware Computing of Tolerance for nucleotide variants), a training-free, parameter-minimal method that infers nucleotide variant tolerability by traversing the mammalian phylogenetic tree and explicitly modeling the evolutionary independence of observed substitutions and their distance from the query species. With only 4 interpretable parameters, no training and no GPU requirement, PHACTn outperforms all evaluated tools on non-coding variants curated from both the ClinVar, and on non-coding variants potentially responsible for selected Mendelian diseases curated from OMIM. Additionally, it achieves state-of-the-art performance on variants within the informative range of alignment-based inference. These results establish that principled probabilistic phylogenetic modeling captures evolutionary constraint signals that large-scale sequence models fail to recover, offering a powerful, accessible, and mechanistically transparent alternative for genome-wide variant effect prediction.

Targeting PFKFB3 Metabolic Gatekeeping of β-Cell Dysfunctional States Enables Durable Glucose Tolerance in Type 2 Diabetes
K. Raval, E. Ilegems, A. C. Title +7
bioRxiv20 Sep 2026PreprintAll rights reservedFull textCell biologybioRxiv
Auditory object identification and localisation training for hyperacusis: preliminary effects on sound tolerance and noise acceptance
Sana Aljoj, Grant Donald Searchfield
Frontiers in human neuroscience24 Oct 2026Peer-reviewedCC BYFull textNeuroscienceEurope PMC
Lysophosphatidylserine Nanoparticles Induce Oral Tolerance Independent of TIM-4 and Canonical PS Processing
E. J. Ovalle, V. Chak, E. Wohlfert +2
bioRxiv24 Sep 2026PreprintCC BYFull textImmunologybioRxiv
A siderophore-independent lanthipeptide carrier enables Trojan-horse delivery of antibacterial cargoes into Gram-negative bacteria.
Boning Wang, Yundan Zheng, Erpeng Guo +9
Chemical science20 Oct 2026Peer-reviewedCC BY-NCFull textMicrobiologyEurope PMC
Culturomics-Guided Precision FMT Enables Stable Antibiotic-Independent Engraftment via Metabolic Niche Targeting
Daniel M. Ferrer, Mark Holton, Minshen Yang +2
bioRxiv25 Oct 2026PreprintCC BY-NCFull textMicrobiologybioRxivPMC

Cite

BibTeX RIS

BibTeX

@misc{yildirim2026phactn,
  title = {{PHACTn enables training-free, context-independent inference of nucleotide variant tolerance across the genome}},
  author = {C. Yildirim and N. Kuru and O. Adebali},
  howpublished = {bioRxiv},
  year = {2026},
  month = sep,
  doi = {10.64898/2026.09.08.750126},
  url = {https://www.biorxiv.org/content/10.64898/2026.09.08.750126v1},
  note = {CC BY-NC-ND},
}

RIS

TY  - UNPB
TI  - PHACTn enables training-free, context-independent inference of nucleotide variant tolerance across the genome
AU  - C. Yildirim
AU  - N. Kuru
AU  - O. Adebali
PY  - 2026
DA  - 2026/09/08
PB  - bioRxiv
DO  - 10.64898/2026.09.08.750126
UR  - https://www.biorxiv.org/content/10.64898/2026.09.08.750126v1
AB  - Accurate prediction of single-nucleotide variant (SNV) tolerability across the entire human genome remains a fundamental challenge in computational genomics, particularly for non-coding regions where the regulatory landscape is vast and poorly understood. Machine learning classifiers suffer from data circularity and demographic bias, while genomic language models demand massive computational resources and offer little biological interpretability. Here, we present PHACTn (Phylogeny-Aware Computing of Tolerance for nucleotide variants), a training-free, parameter-minimal method that infers nucleotide variant tolerability by traversing the mammalian phylogenetic tree and explicitly modeling the evolutionary independence of observed substitutions and their distance from the query species. With only 4 interpretable parameters, no training and no GPU requirement, PHACTn outperforms all evaluated tools on non-coding variants curated from both the ClinVar, and on non-coding variants potentially responsible for selected Mendelian diseases curated from OMIM. Additionally, it achieves state-of-the-art performance on variants within the informative range of alignment-based inference. These results establish that principled probabilistic phylogenetic modeling captures evolutionary constraint signals that large-scale sequence models fail to recover, offering a powerful, accessible, and mechanistically transparent alternative for genome-wide variant effect prediction.
ER  -

Abstract

Related

Cite